From d7e84e8f419ae5bc13f1c1f29d063bf84c21d7db Mon Sep 17 00:00:00 2001 From: bench Date: Sun, 11 Oct 2026 06:26:15 +0900 Subject: [PATCH 1/4] Record the timed operation's allocation counters in every arm --- README.md | 12 ++++ scripts/report.py | 23 +++++-- src/alloc.rs | 83 ++++++++++++++++++++++++++ src/bin/elementwise_fourier.rs | 2 + src/bin/elementwise_gauss2d_patched.rs | 4 ++ src/bin/mpo_mpo_aniso_patched.rs | 4 ++ src/harness.rs | 27 ++++++++- src/lib.rs | 1 + src/record.rs | 12 ++++ 9 files changed, 160 insertions(+), 8 deletions(-) create mode 100644 src/alloc.rs diff --git a/README.md b/README.md index fd6ce3e..5146804 100644 --- a/README.md +++ b/README.md @@ -20,6 +20,18 @@ The terms are combined by deterministic balanced pairwise reduction. Every pair The expensive raw input pair is cached atomically in `.cache/inputs/`. The cache key includes the schema version, pinned tensor4all-rs revision, `N`, `R`, width, aspect range, spacing, polynomial degree, interpolation tolerance, addition tolerance, and seed. Cases 2 and 3 share the same cache entry. Set `BENCH_INPUT_CACHE_REFRESH=1` to rebuild it or `BENCH_INPUT_CACHE_DIR` to move the cache. +## Allocation counters + +Every record carries the allocation calls and the requested bytes of the timed +operation, median over the timed runs. The counters come from a process-wide +`GlobalAlloc` wrapper in `src/alloc.rs`; the benchmark binaries are single +threaded and pinned to one CPU core, so the counts are attributable to the timed +arm. They expose the per-call overhead of the contraction engine, which the wall +time alone hides: at small bond dimensions an arm can allocate two orders of +magnitude more than its tensors hold, and that cost tracks the number of local +updates rather than the tensor size. Records written before the counters existed +omit both fields, and the report generator leaves their columns blank. + ## Patching and accuracy The patch cap is fixed at 128 and has no runtime setting. Gaussian fit arms use relative-L2/SVD truncation. ACI uses its own interpolation residual, which is not identified with an L2 tolerance. Records therefore carry both the internal tolerance metric and a common deterministic holdout sampled relative-L2 error. diff --git a/scripts/report.py b/scripts/report.py index 30ff8e4..8a23d64 100755 --- a/scripts/report.py +++ b/scripts/report.py @@ -13,6 +13,17 @@ } +def alloc_calls(record: dict) -> str: + """Allocation count, blank on records written before the counters existed.""" + value = record.get("n_allocations") + return "" if value is None else str(value) + + +def alloc_mib(record: dict) -> str: + value = record.get("allocated_bytes") + return "" if value is None else f"{value / 2**20:.2f}" + + def metadata(path: Path) -> dict[str, str]: result = {} for line in path.read_text().splitlines(): @@ -58,12 +69,12 @@ def main(profile: Path) -> None: "", "## Case 1: Fourier elementwise", "", - "| input χ | K | arm | time (s) | sampled relative L2 | output χ | parameters |", - "|---:|---:|---|---:|---:|---:|---:|", + "| input χ | K | arm | time (s) | sampled relative L2 | output χ | parameters | allocations | allocated (MiB) |", + "|---:|---:|---|---:|---:|---:|---:|---:|---:|", ] for record in sorted(grouped["elementwise_fourier"], key=lambda item: (item["input_max_bond_dim"], item["params"]["k_max"], item["algorithm"])): lines.append( - f"| {record['input_max_bond_dim']} | {record['params']['k_max']} | {record['algorithm']} | {record['wall_time_median_secs']:.6f} | {record['max_error']:.3e} | {record['output_max_bond_dim']} | {record.get('n_params', '')} |" + f"| {record['input_max_bond_dim']} | {record['params']['k_max']} | {record['algorithm']} | {record['wall_time_median_secs']:.6f} | {record['max_error']:.3e} | {record['output_max_bond_dim']} | {record.get('n_params', '')} | {alloc_calls(record)} | {alloc_mib(record)} |" ) for case, title in [ ("gaussian_elementwise", "Case 2: Gaussian elementwise, global versus patched"), @@ -73,8 +84,8 @@ def main(profile: Path) -> None: "", f"## {title}", "", - "| input χ | raw χ | N | R | arm | time (s) | sampled relative L2 | input patches | input max patch χ | output patches | parameters | speedup |", - "|---:|---:|---:|---:|---|---:|---:|---:|---:|---:|---:|---:|", + "| input χ | raw χ | N | R | arm | time (s) | sampled relative L2 | input patches | input max patch χ | output patches | parameters | allocations | allocated (MiB) | speedup |", + "|---:|---:|---:|---:|---|---:|---:|---:|---:|---:|---:|---:|---:|---:|", ] for record in sorted(grouped[case], key=lambda item: (item["input_max_bond_dim"], item["algorithm"])): params = record["params"] @@ -83,7 +94,7 @@ def main(profile: Path) -> None: input_patch_chi = params.get("input_max_patch_chi", max(params.get("left_input_max_patch_chi", 0), params.get("right_input_max_patch_chi", 0))) speedup = params.get("speedup_vs_global", 1.0) lines.append( - f"| {record['input_max_bond_dim']} | {raw_chi} | {params['n_gauss']} | {params['r']} | {record['algorithm']} | {record['wall_time_median_secs']:.6f} | {record['max_error']:.3e} | {input_patches} | {input_patch_chi} | {record.get('n_patches', 1)} | {record.get('n_params', '')} | {speedup:.3f} |" + f"| {record['input_max_bond_dim']} | {raw_chi} | {params['n_gauss']} | {params['r']} | {record['algorithm']} | {record['wall_time_median_secs']:.6f} | {record['max_error']:.3e} | {input_patches} | {input_patch_chi} | {record.get('n_patches', 1)} | {record.get('n_params', '')} | {alloc_calls(record)} | {alloc_mib(record)} | {speedup:.3f} |" ) (profile / "report.md").write_text("\n".join(lines) + "\n") diff --git a/src/alloc.rs b/src/alloc.rs new file mode 100644 index 0000000..a1b5548 --- /dev/null +++ b/src/alloc.rs @@ -0,0 +1,83 @@ +//! Process-wide allocation counters for the timed arms. +//! +//! The maintained cases run small tensor networks, where per-call heap traffic +//! can outweigh the floating-point work: a variational fit at twenty sites with +//! bond dimension below thirty allocates two orders of magnitude more than the +//! network holds, and the count depends on the number of local updates rather +//! than on the tensor size. Recording allocations next to the wall time keeps +//! that overhead visible, so an upstream change to the contraction engine shows +//! up as a number instead of as an unexplained timing shift. +//! +//! The counters are only read around the timed region, so the two relaxed atomic +//! adds per allocation stay far below the timing noise. + +use std::alloc::{GlobalAlloc, Layout, System}; +use std::sync::atomic::{AtomicU64, Ordering}; + +pub struct CountingAllocator; + +static ALLOC_CALLS: AtomicU64 = AtomicU64::new(0); +static ALLOC_BYTES: AtomicU64 = AtomicU64::new(0); + +// Only `alloc` and `dealloc` are overridden. The trait's default `realloc` and +// `alloc_zeroed` both route through `self.alloc`, so every allocation event is +// counted exactly once without duplicating the system calls here. +unsafe impl GlobalAlloc for CountingAllocator { + unsafe fn alloc(&self, layout: Layout) -> *mut u8 { + ALLOC_CALLS.fetch_add(1, Ordering::Relaxed); + ALLOC_BYTES.fetch_add(layout.size() as u64, Ordering::Relaxed); + unsafe { System.alloc(layout) } + } + + unsafe fn dealloc(&self, ptr: *mut u8, layout: Layout) { + unsafe { System.dealloc(ptr, layout) } + } +} + +#[global_allocator] +static ALLOCATOR: CountingAllocator = CountingAllocator; + +/// Allocation counters at one instant, in allocation calls and requested bytes. +#[derive(Clone, Copy, Debug, Default, PartialEq, Eq)] +pub struct AllocSnapshot { + pub calls: u64, + pub bytes: u64, +} + +pub fn snapshot() -> AllocSnapshot { + AllocSnapshot { + calls: ALLOC_CALLS.load(Ordering::Relaxed), + bytes: ALLOC_BYTES.load(Ordering::Relaxed), + } +} + +/// Counters accumulated since `earlier`. +pub fn delta(earlier: AllocSnapshot) -> AllocSnapshot { + let now = snapshot(); + AllocSnapshot { + calls: now.calls - earlier.calls, + bytes: now.bytes - earlier.bytes, + } +} + +#[cfg(test)] +mod tests { + use super::*; + + #[test] + fn delta_counts_a_known_allocation() { + let before = snapshot(); + let payload: Vec = vec![7u8; 4096]; + let after = delta(before); + assert!( + after.calls >= 1, + "allocating a 4 KiB vector must be counted" + ); + assert!( + after.bytes >= 4096, + "the requested bytes must cover the vector payload, got {}", + after.bytes + ); + assert_eq!(payload.iter().filter(|&&b| b == 7).count(), 4096); + } +} diff --git a/src/bin/elementwise_fourier.rs b/src/bin/elementwise_fourier.rs index 857810b..e600903 100644 --- a/src/bin/elementwise_fourier.rs +++ b/src/bin/elementwise_fourier.rs @@ -138,6 +138,8 @@ fn main() -> anyhow::Result<()> { max_patch_bond: None, rtol: None, input_build_secs: None, + n_allocations: Some(timing.alloc_calls_median), + allocated_bytes: Some(timing.alloc_bytes_median), }; write_record( &out_dir, diff --git a/src/bin/elementwise_gauss2d_patched.rs b/src/bin/elementwise_gauss2d_patched.rs index 7fcb3de..9120f9f 100644 --- a/src/bin/elementwise_gauss2d_patched.rs +++ b/src/bin/elementwise_gauss2d_patched.rs @@ -395,6 +395,8 @@ fn write_global_record( max_patch_bond: None, rtol: (tolerance_metric == "relative_l2_svd").then_some(INPUT_L2_RTOL), input_build_secs: Some(input.build.as_secs_f64()), + n_allocations: Some(timing.alloc_calls_median), + allocated_bytes: Some(timing.alloc_bytes_median), }, ) } @@ -456,6 +458,8 @@ fn write_patched_record( max_patch_bond: Some(max_patch_bond(output)), rtol: Some(INPUT_L2_RTOL), input_build_secs: Some(input.build.as_secs_f64()), + n_allocations: Some(timing.alloc_calls_median), + allocated_bytes: Some(timing.alloc_bytes_median), }, ) } diff --git a/src/bin/mpo_mpo_aniso_patched.rs b/src/bin/mpo_mpo_aniso_patched.rs index 86ba0b9..1ee04f8 100644 --- a/src/bin/mpo_mpo_aniso_patched.rs +++ b/src/bin/mpo_mpo_aniso_patched.rs @@ -150,6 +150,8 @@ fn run_point( max_patch_bond: None, rtol: Some(INPUT_L2_RTOL), input_build_secs: Some(input.build.as_secs_f64()), + n_allocations: Some(global_timing.alloc_calls_median), + allocated_bytes: Some(global_timing.alloc_bytes_median), }, )?; @@ -191,6 +193,8 @@ fn run_point( seed, tolerance: INPUT_L2_RTOL, wall_time_median_secs: patched_timing.median_secs, + n_allocations: Some(patched_timing.alloc_calls_median), + allocated_bytes: Some(patched_timing.alloc_bytes_median), wall_times_secs: patched_timing.runs_secs, max_error: patched_error, input_max_bond_dim: input_chi, diff --git a/src/harness.rs b/src/harness.rs index 7dd4774..704bf63 100644 --- a/src/harness.rs +++ b/src/harness.rs @@ -6,6 +6,10 @@ use std::time::Instant; pub struct Timing { pub median_secs: f64, pub runs_secs: Vec, + /// Heap allocation calls made by the timed operation, median over the runs. + pub alloc_calls_median: u64, + /// Bytes requested by those calls, median over the runs. + pub alloc_bytes_median: u64, } /// Run `f` `warmups` times untimed, then `repeats` times timed. @@ -16,11 +20,17 @@ pub fn time_median(warmups: usize, repeats: usize, mut f: impl FnMut() -> R) std::hint::black_box(f()); } let mut runs_secs = Vec::with_capacity(repeats); + let mut alloc_calls = Vec::with_capacity(repeats); + let mut alloc_bytes = Vec::with_capacity(repeats); let mut last = None; for _ in 0..repeats { + let before = crate::alloc::snapshot(); let t0 = Instant::now(); last = Some(std::hint::black_box(f())); runs_secs.push(t0.elapsed().as_secs_f64()); + let allocated = crate::alloc::delta(before); + alloc_calls.push(allocated.calls); + alloc_bytes.push(allocated.bytes); } let mut sorted = runs_secs.clone(); sorted.sort_by(|a, b| a.partial_cmp(b).unwrap()); @@ -30,10 +40,18 @@ pub fn time_median(warmups: usize, repeats: usize, mut f: impl FnMut() -> R) Timing { median_secs, runs_secs, + alloc_calls_median: median_u64(&alloc_calls), + alloc_bytes_median: median_u64(&alloc_bytes), }, ) } +fn median_u64(values: &[u64]) -> u64 { + let mut sorted = values.to_vec(); + sorted.sort_unstable(); + sorted[sorted.len() / 2] +} + /// `n` random grid indices in [0, 2^r), deterministic in `seed`. pub fn sample_grid_indices(r: usize, n: usize, seed: u64) -> Vec { let mut rng = ChaCha8Rng::seed_from_u64(seed); @@ -51,10 +69,15 @@ mod tests { #[test] fn time_median_returns_result_and_positive_times() { - let (val, timing) = time_median(1, 3, || 40 + 2); - assert_eq!(val, 42); + let (val, timing) = time_median(1, 3, || vec![40u8, 2]); + assert_eq!(val, vec![40u8, 2]); assert_eq!(timing.runs_secs.len(), 3); assert!(timing.median_secs >= 0.0); + assert!( + timing.alloc_calls_median >= 1, + "the timed closure allocates" + ); + assert!(timing.alloc_bytes_median >= 2); } #[test] diff --git a/src/lib.rs b/src/lib.rs index 1558e7a..5af0f1d 100644 --- a/src/lib.rs +++ b/src/lib.rs @@ -1,3 +1,4 @@ +pub mod alloc; pub mod elementwise; pub mod fourier; pub mod gaussian; diff --git a/src/record.rs b/src/record.rs index b9d2099..152a941 100644 --- a/src/record.rs +++ b/src/record.rs @@ -45,6 +45,14 @@ pub struct RunRecord { /// inputs are shared by every arm of the instance. #[serde(skip_serializing_if = "Option::is_none")] pub input_build_secs: Option, + /// Heap allocation calls made by the timed operation, median over the timed + /// runs. Present on every arm measured by the current harness; absent from + /// records written before the counters existed. + #[serde(skip_serializing_if = "Option::is_none")] + pub n_allocations: Option, + /// Bytes requested by those allocation calls, median over the timed runs. + #[serde(skip_serializing_if = "Option::is_none")] + pub allocated_bytes: Option, } pub fn write_record(dir: &Path, name: &str, record: &RunRecord) -> anyhow::Result<()> { @@ -78,12 +86,16 @@ mod tests { max_patch_bond: None, rtol: None, input_build_secs: None, + n_allocations: Some(1234), + allocated_bytes: Some(5678), }; let s = serde_json::to_string(&rec).unwrap(); let v: serde_json::Value = serde_json::from_str(&s).unwrap(); assert_eq!(v["schema_version"], 1); assert_eq!(v["params"]["k_max"], 8); assert_eq!(v["n_params"], 120); + assert_eq!(v["n_allocations"], 1234); + assert_eq!(v["allocated_bytes"], 5678); // The unset optional fields are omitted rather than written as null, so // a record only ever carries the fields its case actually measured. assert!( From b8fcf4848e7afda81029c8f9d48900ac58b2fa1a Mon Sep 17 00:00:00 2001 From: bench Date: Sun, 11 Oct 2026 07:25:21 +0900 Subject: [PATCH 2/4] Track tensor4all-rs a515738 and migrate the TreeTN patch cutoff --- Cargo.lock | 236 ++++++++++++++++++++++++++++++++------------- Cargo.toml | 18 ++-- scripts/run_all.sh | 2 +- src/patched_mpo.rs | 8 +- 4 files changed, 185 insertions(+), 79 deletions(-) diff --git a/Cargo.lock b/Cargo.lock index 47e3678..2c351fa 100644 --- a/Cargo.lock +++ b/Cargo.lock @@ -11,12 +11,6 @@ dependencies = [ "memchr", ] -[[package]] -name = "allocator-api2" -version = "0.2.21" -source = "registry+https://github.com/rust-lang/crates.io-index" -checksum = "683d7910e743518b0e34f1186f92494becacb047c7b6bf616c96772180fef923" - [[package]] name = "anyhow" version = "1.0.104" @@ -104,6 +98,18 @@ name = "computegraph" version = "0.1.0" source = "git+https://github.com/tensor4all/computegraph-rs.git?rev=691def2d82fd4367b397d61209449f68e82050b7#691def2d82fd4367b397d61209449f68e82050b7" +[[package]] +name = "cpueinsum" +version = "0.1.0" +source = "git+https://github.com/tensor4all/cpueinsum-rs.git?rev=e87d0d1bfab3f1353e5e57fe91ec0442ac99a4b2#e87d0d1bfab3f1353e5e57fe91ec0442ac99a4b2" +dependencies = [ + "num-complex", + "strided-view", + "thiserror 1.0.69", + "tprims-contract", + "tprims-exec", +] + [[package]] name = "crossbeam" version = "0.8.4" @@ -327,9 +333,9 @@ checksum = "1d674e81391d1e1ab681a28d99df07927c6d4aa5b027d7da16ba32d1d21ecd99" [[package]] name = "foldhash" -version = "0.1.5" +version = "0.2.0" source = "registry+https://github.com/rust-lang/crates.io-index" -checksum = "d9c4f5dac5e15c24eb999c26181a6ca40b39fe946cbe4c263c7209467bc83af2" +checksum = "77ce24cb58228fbb8aa041425bb1050850ac19177686ea6e0f41a70416f56fdb" [[package]] name = "gemm" @@ -498,21 +504,13 @@ dependencies = [ [[package]] name = "hashbrown" -version = "0.15.5" +version = "0.17.1" source = "registry+https://github.com/rust-lang/crates.io-index" -checksum = "9229cfe53dfd69f0609a49f65461bd93001ea1ef889cd5529dd176593f5338a1" +checksum = "ed5909b6e89a2db4456e54cd5f673791d7eca6732202bbf2a9cc504fe2f9b84a" dependencies = [ - "allocator-api2", - "equivalent", "foldhash", ] -[[package]] -name = "hashbrown" -version = "0.17.1" -source = "registry+https://github.com/rust-lang/crates.io-index" -checksum = "ed5909b6e89a2db4456e54cd5f673791d7eca6732202bbf2a9cc504fe2f9b84a" - [[package]] name = "hdf5-metno" version = "0.12.6" @@ -589,7 +587,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index" checksum = "d466e9454f08e4a911e14806c24e16fba1b4c121d1ea474396f396069cf949d9" dependencies = [ "equivalent", - "hashbrown 0.17.1", + "hashbrown", ] [[package]] @@ -667,11 +665,11 @@ dependencies = [ [[package]] name = "lru" -version = "0.12.5" +version = "0.18.5" source = "registry+https://github.com/rust-lang/crates.io-index" -checksum = "234cf4f4a04dc1f57e24b96cc0cd600cf2af460d4161ac5ecdd0af8e1f3b2a38" +checksum = "ef9ac18847474e638e3702b76c65d4eb93428471a74778ef0f1be711717f89b5" dependencies = [ - "hashbrown 0.15.5", + "hashbrown", ] [[package]] @@ -1257,9 +1255,9 @@ dependencies = [ ] [[package]] -name = "strided-kernel" -version = "0.4.0" -source = "git+https://github.com/tensor4all/strided-rs.git?rev=b29e7601ba090aa5eafc65b9bde5d9450282e0d8#b29e7601ba090aa5eafc65b9bde5d9450282e0d8" +name = "strided-basic" +version = "0.4.6" +source = "git+https://github.com/tensor4all/strided-rs.git?rev=12ff2de906f4d894a4329f7dbf195b0f874f846e#12ff2de906f4d894a4329f7dbf195b0f874f846e" dependencies = [ "num-complex", "num-traits", @@ -1270,10 +1268,29 @@ dependencies = [ "strided-view", ] +[[package]] +name = "strided-fused" +version = "0.4.6" +source = "git+https://github.com/tensor4all/strided-rs.git?rev=12ff2de906f4d894a4329f7dbf195b0f874f846e#12ff2de906f4d894a4329f7dbf195b0f874f846e" +dependencies = [ + "num-complex", + "strided-basic", +] + +[[package]] +name = "strided-kernel" +version = "0.4.6" +source = "git+https://github.com/tensor4all/strided-rs.git?rev=12ff2de906f4d894a4329f7dbf195b0f874f846e#12ff2de906f4d894a4329f7dbf195b0f874f846e" +dependencies = [ + "num-complex", + "strided-basic", + "strided-fused", +] + [[package]] name = "strided-perm" -version = "0.4.0" -source = "git+https://github.com/tensor4all/strided-rs.git?rev=b29e7601ba090aa5eafc65b9bde5d9450282e0d8#b29e7601ba090aa5eafc65b9bde5d9450282e0d8" +version = "0.4.6" +source = "git+https://github.com/tensor4all/strided-rs.git?rev=12ff2de906f4d894a4329f7dbf195b0f874f846e#12ff2de906f4d894a4329f7dbf195b0f874f846e" dependencies = [ "rayon", "strided-view", @@ -1282,8 +1299,8 @@ dependencies = [ [[package]] name = "strided-traits" -version = "0.4.0" -source = "git+https://github.com/tensor4all/strided-rs.git?rev=b29e7601ba090aa5eafc65b9bde5d9450282e0d8#b29e7601ba090aa5eafc65b9bde5d9450282e0d8" +version = "0.4.6" +source = "git+https://github.com/tensor4all/strided-rs.git?rev=12ff2de906f4d894a4329f7dbf195b0f874f846e#12ff2de906f4d894a4329f7dbf195b0f874f846e" dependencies = [ "num-complex", "num-traits", @@ -1291,8 +1308,8 @@ dependencies = [ [[package]] name = "strided-view" -version = "0.4.0" -source = "git+https://github.com/tensor4all/strided-rs.git?rev=b29e7601ba090aa5eafc65b9bde5d9450282e0d8#b29e7601ba090aa5eafc65b9bde5d9450282e0d8" +version = "0.4.6" +source = "git+https://github.com/tensor4all/strided-rs.git?rev=12ff2de906f4d894a4329f7dbf195b0f874f846e#12ff2de906f4d894a4329f7dbf195b0f874f846e" dependencies = [ "num-complex", "num-traits", @@ -1371,13 +1388,14 @@ dependencies = [ [[package]] name = "tenferro-ad" -version = "0.3.0" -source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=b5a106be3133979d78832a0ca3f4d6b57613b3d7#b5a106be3133979d78832a0ca3f4d6b57613b3d7" +version = "0.7.1" +source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=ff94aeded9cc98d6c49889c8ef6fca365763d93a#ff94aeded9cc98d6c49889c8ef6fca365763d93a" dependencies = [ "computegraph", "lru", "num-complex", "num-traits", + "rayon", "tenferro-cpu", "tenferro-internal-ops", "tenferro-runtime", @@ -1387,15 +1405,17 @@ dependencies = [ [[package]] name = "tenferro-core-ops" -version = "0.3.0" -source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=b5a106be3133979d78832a0ca3f4d6b57613b3d7#b5a106be3133979d78832a0ca3f4d6b57613b3d7" +version = "0.7.1" +source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=ff94aeded9cc98d6c49889c8ef6fca365763d93a#ff94aeded9cc98d6c49889c8ef6fca365763d93a" [[package]] name = "tenferro-cpu" -version = "0.3.0" -source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=b5a106be3133979d78832a0ca3f4d6b57613b3d7#b5a106be3133979d78832a0ca3f4d6b57613b3d7" +version = "0.7.1" +source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=ff94aeded9cc98d6c49889c8ef6fca365763d93a#ff94aeded9cc98d6c49889c8ef6fca365763d93a" dependencies = [ + "cpueinsum", "faer", + "libm", "lru", "num-complex", "num-traits", @@ -1403,23 +1423,53 @@ dependencies = [ "smallvec", "strided-kernel", "tenferro-core-ops", + "tenferro-cpu-basic", + "tenferro-cpu-fused", "tenferro-internal-cpu-kernels", "tenferro-runtime", "tenferro-tensor", "thiserror 2.0.20", ] +[[package]] +name = "tenferro-cpu-basic" +version = "0.7.1" +source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=ff94aeded9cc98d6c49889c8ef6fca365763d93a#ff94aeded9cc98d6c49889c8ef6fca365763d93a" +dependencies = [ + "num-complex", + "num-traits", + "strided-basic", + "tenferro-tensor", + "thiserror 2.0.20", +] + +[[package]] +name = "tenferro-cpu-fused" +version = "0.7.1" +source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=ff94aeded9cc98d6c49889c8ef6fca365763d93a#ff94aeded9cc98d6c49889c8ef6fca365763d93a" +dependencies = [ + "num-complex", + "num-traits", + "strided-basic", + "strided-fused", + "tenferro-cpu-basic", + "tenferro-tensor", + "thiserror 2.0.20", +] + [[package]] name = "tenferro-einsum" -version = "0.3.0" -source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=b5a106be3133979d78832a0ca3f4d6b57613b3d7#b5a106be3133979d78832a0ca3f4d6b57613b3d7" +version = "0.7.1" +source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=ff94aeded9cc98d6c49889c8ef6fca365763d93a#ff94aeded9cc98d6c49889c8ef6fca365763d93a" dependencies = [ "computegraph", + "cpueinsum", "lru", "omeco", "smallvec", "tenferro-ad", "tenferro-cpu", + "tenferro-cpu-basic", "tenferro-internal-extension-macros", "tenferro-internal-ops", "tenferro-runtime", @@ -1429,20 +1479,23 @@ dependencies = [ [[package]] name = "tenferro-internal-cpu-kernels" -version = "0.3.0" -source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=b5a106be3133979d78832a0ca3f4d6b57613b3d7#b5a106be3133979d78832a0ca3f4d6b57613b3d7" +version = "0.7.1" +source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=ff94aeded9cc98d6c49889c8ef6fca365763d93a#ff94aeded9cc98d6c49889c8ef6fca365763d93a" dependencies = [ "num-complex", "num-traits", + "smallvec", "strided-kernel", + "tenferro-cpu-basic", "tenferro-tensor", + "tenferro-tensor-core", "thiserror 2.0.20", ] [[package]] name = "tenferro-internal-extension-macros" -version = "0.3.0" -source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=b5a106be3133979d78832a0ca3f4d6b57613b3d7#b5a106be3133979d78832a0ca3f4d6b57613b3d7" +version = "0.7.1" +source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=ff94aeded9cc98d6c49889c8ef6fca365763d93a#ff94aeded9cc98d6c49889c8ef6fca365763d93a" dependencies = [ "proc-macro2", "quote", @@ -1451,8 +1504,8 @@ dependencies = [ [[package]] name = "tenferro-internal-ops" -version = "0.3.0" -source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=b5a106be3133979d78832a0ca3f4d6b57613b3d7#b5a106be3133979d78832a0ca3f4d6b57613b3d7" +version = "0.7.1" +source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=ff94aeded9cc98d6c49889c8ef6fca365763d93a#ff94aeded9cc98d6c49889c8ef6fca365763d93a" dependencies = [ "computegraph", "num-complex", @@ -1464,13 +1517,15 @@ dependencies = [ [[package]] name = "tenferro-linalg" -version = "0.3.0" -source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=b5a106be3133979d78832a0ca3f4d6b57613b3d7#b5a106be3133979d78832a0ca3f4d6b57613b3d7" +version = "0.7.1" +source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=ff94aeded9cc98d6c49889c8ef6fca365763d93a#ff94aeded9cc98d6c49889c8ef6fca365763d93a" dependencies = [ "computegraph", "faer", "num-complex", "num-traits", + "rayon", + "strided-view", "tenferro-ad", "tenferro-cpu", "tenferro-internal-extension-macros", @@ -1478,12 +1533,13 @@ dependencies = [ "tenferro-runtime", "tenferro-tensor", "thiserror 2.0.20", + "tlinalg", ] [[package]] name = "tenferro-runtime" -version = "0.3.0" -source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=b5a106be3133979d78832a0ca3f4d6b57613b3d7#b5a106be3133979d78832a0ca3f4d6b57613b3d7" +version = "0.7.1" +source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=ff94aeded9cc98d6c49889c8ef6fca365763d93a#ff94aeded9cc98d6c49889c8ef6fca365763d93a" dependencies = [ "computegraph", "lru", @@ -1491,6 +1547,7 @@ dependencies = [ "num-traits", "smallvec", "tenferro-core-ops", + "tenferro-internal-extension-macros", "tenferro-internal-ops", "tenferro-tensor", "thiserror 2.0.20", @@ -1498,13 +1555,12 @@ dependencies = [ [[package]] name = "tenferro-tensor" -version = "0.3.0" -source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=b5a106be3133979d78832a0ca3f4d6b57613b3d7#b5a106be3133979d78832a0ca3f4d6b57613b3d7" +version = "0.7.1" +source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=ff94aeded9cc98d6c49889c8ef6fca365763d93a#ff94aeded9cc98d6c49889c8ef6fca365763d93a" dependencies = [ "num-complex", "num-traits", "smallvec", - "strided-kernel", "tenferro-core-ops", "tenferro-tensor-core", "thiserror 2.0.20", @@ -1512,8 +1568,8 @@ dependencies = [ [[package]] name = "tenferro-tensor-core" -version = "0.3.0" -source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=b5a106be3133979d78832a0ca3f4d6b57613b3d7#b5a106be3133979d78832a0ca3f4d6b57613b3d7" +version = "0.7.1" +source = "git+https://github.com/tensor4all/tenferro-rs.git?rev=ff94aeded9cc98d6c49889c8ef6fca365763d93a#ff94aeded9cc98d6c49889c8ef6fca365763d93a" dependencies = [ "num-complex", "smallvec", @@ -1523,7 +1579,7 @@ dependencies = [ [[package]] name = "tensor4all-aci" version = "0.2.0" -source = "git+https://github.com/tensor4all/tensor4all-rs?rev=9e9aedaebe0d3918b34dd399ff0981e337f3835b#9e9aedaebe0d3918b34dd399ff0981e337f3835b" +source = "git+https://github.com/tensor4all/tensor4all-rs?rev=a5157385f58048355a28a6a003d4cb63d3b1349b#a5157385f58048355a28a6a003d4cb63d3b1349b" dependencies = [ "num-complex", "rand", @@ -1538,7 +1594,7 @@ dependencies = [ [[package]] name = "tensor4all-core" version = "0.2.0" -source = "git+https://github.com/tensor4all/tensor4all-rs?rev=9e9aedaebe0d3918b34dd399ff0981e337f3835b#9e9aedaebe0d3918b34dd399ff0981e337f3835b" +source = "git+https://github.com/tensor4all/tensor4all-rs?rev=a5157385f58048355a28a6a003d4cb63d3b1349b#a5157385f58048355a28a6a003d4cb63d3b1349b" dependencies = [ "anyhow", "bnum", @@ -1563,7 +1619,7 @@ dependencies = [ [[package]] name = "tensor4all-hdf5" version = "0.2.0" -source = "git+https://github.com/tensor4all/tensor4all-rs?rev=9e9aedaebe0d3918b34dd399ff0981e337f3835b#9e9aedaebe0d3918b34dd399ff0981e337f3835b" +source = "git+https://github.com/tensor4all/tensor4all-rs?rev=a5157385f58048355a28a6a003d4cb63d3b1349b#a5157385f58048355a28a6a003d4cb63d3b1349b" dependencies = [ "anyhow", "hdf5-metno", @@ -1579,7 +1635,7 @@ dependencies = [ [[package]] name = "tensor4all-interpolativeqtt" version = "0.2.0" -source = "git+https://github.com/tensor4all/tensor4all-rs?rev=9e9aedaebe0d3918b34dd399ff0981e337f3835b#9e9aedaebe0d3918b34dd399ff0981e337f3835b" +source = "git+https://github.com/tensor4all/tensor4all-rs?rev=a5157385f58048355a28a6a003d4cb63d3b1349b#a5157385f58048355a28a6a003d4cb63d3b1349b" dependencies = [ "tensor4all-simplett", "tensor4all-tensorbackend", @@ -1589,13 +1645,14 @@ dependencies = [ [[package]] name = "tensor4all-itensorlike" version = "0.2.0" -source = "git+https://github.com/tensor4all/tensor4all-rs?rev=9e9aedaebe0d3918b34dd399ff0981e337f3835b#9e9aedaebe0d3918b34dd399ff0981e337f3835b" +source = "git+https://github.com/tensor4all/tensor4all-rs?rev=a5157385f58048355a28a6a003d4cb63d3b1349b#a5157385f58048355a28a6a003d4cb63d3b1349b" dependencies = [ "anyhow", "num-complex", "petgraph", "rand", "tensor4all-core", + "tensor4all-tensorbackend", "tensor4all-treetn", "thiserror 2.0.20", ] @@ -1603,11 +1660,10 @@ dependencies = [ [[package]] name = "tensor4all-partitionedtreetn" version = "0.2.0" -source = "git+https://github.com/tensor4all/tensor4all-rs?rev=9e9aedaebe0d3918b34dd399ff0981e337f3835b#9e9aedaebe0d3918b34dd399ff0981e337f3835b" +source = "git+https://github.com/tensor4all/tensor4all-rs?rev=a5157385f58048355a28a6a003d4cb63d3b1349b#a5157385f58048355a28a6a003d4cb63d3b1349b" dependencies = [ "anyhow", "tensor4all-core", - "tensor4all-tensorbackend", "tensor4all-treetn", "thiserror 2.0.20", ] @@ -1615,12 +1671,13 @@ dependencies = [ [[package]] name = "tensor4all-partitionedtt" version = "0.2.0" -source = "git+https://github.com/tensor4all/tensor4all-rs?rev=9e9aedaebe0d3918b34dd399ff0981e337f3835b#9e9aedaebe0d3918b34dd399ff0981e337f3835b" +source = "git+https://github.com/tensor4all/tensor4all-rs?rev=a5157385f58048355a28a6a003d4cb63d3b1349b#a5157385f58048355a28a6a003d4cb63d3b1349b" dependencies = [ "anyhow", "num-complex", "num-traits", "rand", + "rand_chacha", "tensor4all-core", "tensor4all-itensorlike", "tensor4all-simplett", @@ -1633,7 +1690,7 @@ dependencies = [ [[package]] name = "tensor4all-simplett" version = "0.2.0" -source = "git+https://github.com/tensor4all/tensor4all-rs?rev=9e9aedaebe0d3918b34dd399ff0981e337f3835b#9e9aedaebe0d3918b34dd399ff0981e337f3835b" +source = "git+https://github.com/tensor4all/tensor4all-rs?rev=a5157385f58048355a28a6a003d4cb63d3b1349b#a5157385f58048355a28a6a003d4cb63d3b1349b" dependencies = [ "anyhow", "bnum", @@ -1650,7 +1707,7 @@ dependencies = [ [[package]] name = "tensor4all-tensorbackend" version = "0.2.0" -source = "git+https://github.com/tensor4all/tensor4all-rs?rev=9e9aedaebe0d3918b34dd399ff0981e337f3835b#9e9aedaebe0d3918b34dd399ff0981e337f3835b" +source = "git+https://github.com/tensor4all/tensor4all-rs?rev=a5157385f58048355a28a6a003d4cb63d3b1349b#a5157385f58048355a28a6a003d4cb63d3b1349b" dependencies = [ "anyhow", "num-complex", @@ -1658,6 +1715,7 @@ dependencies = [ "omeco", "rand", "rand_distr", + "rayon", "tenferro-ad", "tenferro-cpu", "tenferro-einsum", @@ -1670,12 +1728,13 @@ dependencies = [ [[package]] name = "tensor4all-tensorci" version = "0.2.0" -source = "git+https://github.com/tensor4all/tensor4all-rs?rev=9e9aedaebe0d3918b34dd399ff0981e337f3835b#9e9aedaebe0d3918b34dd399ff0981e337f3835b" +source = "git+https://github.com/tensor4all/tensor4all-rs?rev=a5157385f58048355a28a6a003d4cb63d3b1349b#a5157385f58048355a28a6a003d4cb63d3b1349b" dependencies = [ "anyhow", "num-complex", "num-traits", "rand", + "rand_chacha", "tensor4all-core", "tensor4all-simplett", "tensor4all-tensorbackend", @@ -1685,13 +1744,14 @@ dependencies = [ [[package]] name = "tensor4all-treetn" version = "0.2.0" -source = "git+https://github.com/tensor4all/tensor4all-rs?rev=9e9aedaebe0d3918b34dd399ff0981e337f3835b#9e9aedaebe0d3918b34dd399ff0981e337f3835b" +source = "git+https://github.com/tensor4all/tensor4all-rs?rev=a5157385f58048355a28a6a003d4cb63d3b1349b#a5157385f58048355a28a6a003d4cb63d3b1349b" dependencies = [ "anyhow", "dyn-clone", "num-complex", "petgraph", "rand", + "rand_chacha", "tensor4all-core", "tensor4all-simplett", "tensor4all-tensorbackend", @@ -1747,6 +1807,18 @@ dependencies = [ "cfg-if", ] +[[package]] +name = "tlinalg" +version = "0.1.0" +source = "git+https://github.com/tensor4all/tlinalg-rs.git?rev=e2e6c920150b1383338fef581abeb5affc8997c4#e2e6c920150b1383338fef581abeb5affc8997c4" +dependencies = [ + "faer", + "num-complex", + "rayon", + "strided-view", + "thiserror 2.0.20", +] + [[package]] name = "toml_datetime" version = "1.1.1+spec-1.1.0" @@ -1777,6 +1849,38 @@ dependencies = [ "winnow", ] +[[package]] +name = "tprims-contract" +version = "0.1.0" +source = "git+https://github.com/tensor4all/tprims-rs.git?rev=c8b99d608879dfe09b8f8daaf5a1d48aa0b94310#c8b99d608879dfe09b8f8daaf5a1d48aa0b94310" +dependencies = [ + "faer", + "num-complex", + "strided-basic", + "strided-view", + "thiserror 1.0.69", + "tprims-exec", + "tprims-kernel", +] + +[[package]] +name = "tprims-exec" +version = "0.1.0" +source = "git+https://github.com/tensor4all/tprims-rs.git?rev=c8b99d608879dfe09b8f8daaf5a1d48aa0b94310#c8b99d608879dfe09b8f8daaf5a1d48aa0b94310" +dependencies = [ + "rayon", + "strided-basic", + "thiserror 1.0.69", +] + +[[package]] +name = "tprims-kernel" +version = "0.1.0" +source = "git+https://github.com/tensor4all/tprims-rs.git?rev=c8b99d608879dfe09b8f8daaf5a1d48aa0b94310#c8b99d608879dfe09b8f8daaf5a1d48aa0b94310" +dependencies = [ + "num-complex", +] + [[package]] name = "tracing" version = "0.1.44" diff --git a/Cargo.toml b/Cargo.toml index 1b463b9..a42a2df 100644 --- a/Cargo.toml +++ b/Cargo.toml @@ -6,15 +6,15 @@ license = "MIT" publish = false [dependencies] -tensor4all-simplett = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "9e9aedaebe0d3918b34dd399ff0981e337f3835b", package = "tensor4all-simplett" } -tensor4all-aci = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "9e9aedaebe0d3918b34dd399ff0981e337f3835b", package = "tensor4all-aci" } -tensor4all-treetn = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "9e9aedaebe0d3918b34dd399ff0981e337f3835b", package = "tensor4all-treetn" } -tensor4all-itensorlike = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "9e9aedaebe0d3918b34dd399ff0981e337f3835b", package = "tensor4all-itensorlike" } -tensor4all-partitionedtt = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "9e9aedaebe0d3918b34dd399ff0981e337f3835b", package = "tensor4all-partitionedtt" } -tensor4all-partitionedtreetn = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "9e9aedaebe0d3918b34dd399ff0981e337f3835b", package = "tensor4all-partitionedtreetn" } -tensor4all-interpolativeqtt = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "9e9aedaebe0d3918b34dd399ff0981e337f3835b", package = "tensor4all-interpolativeqtt" } -tensor4all-hdf5 = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "9e9aedaebe0d3918b34dd399ff0981e337f3835b", package = "tensor4all-hdf5" } -tensor4all-core = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "9e9aedaebe0d3918b34dd399ff0981e337f3835b", package = "tensor4all-core" } +tensor4all-simplett = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "a5157385f58048355a28a6a003d4cb63d3b1349b", package = "tensor4all-simplett" } +tensor4all-aci = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "a5157385f58048355a28a6a003d4cb63d3b1349b", package = "tensor4all-aci" } +tensor4all-treetn = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "a5157385f58048355a28a6a003d4cb63d3b1349b", package = "tensor4all-treetn" } +tensor4all-itensorlike = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "a5157385f58048355a28a6a003d4cb63d3b1349b", package = "tensor4all-itensorlike" } +tensor4all-partitionedtt = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "a5157385f58048355a28a6a003d4cb63d3b1349b", package = "tensor4all-partitionedtt" } +tensor4all-partitionedtreetn = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "a5157385f58048355a28a6a003d4cb63d3b1349b", package = "tensor4all-partitionedtreetn" } +tensor4all-interpolativeqtt = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "a5157385f58048355a28a6a003d4cb63d3b1349b", package = "tensor4all-interpolativeqtt" } +tensor4all-hdf5 = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "a5157385f58048355a28a6a003d4cb63d3b1349b", package = "tensor4all-hdf5" } +tensor4all-core = { git = "https://github.com/tensor4all/tensor4all-rs", rev = "a5157385f58048355a28a6a003d4cb63d3b1349b", package = "tensor4all-core" } num-complex = "0.4" rand = "0.9" rand_chacha = "0.9" diff --git a/scripts/run_all.sh b/scripts/run_all.sh index cab0ef1..5373181 100755 --- a/scripts/run_all.sh +++ b/scripts/run_all.sh @@ -44,7 +44,7 @@ os: $(uname -s) chip: ${chip:-unknown} memory_gb: ${memory_gb:-unknown} repo_rev: $repo_rev -tensor4all_rs_rev: 9e9aedaebe0d3918b34dd399ff0981e337f3835b +tensor4all_rs_rev: a5157385f58048355a28a6a003d4cb63d3b1349b threads: 1 cpu_affinity: $core EOF diff --git a/src/patched_mpo.rs b/src/patched_mpo.rs index 741bd18..dd98fe1 100644 --- a/src/patched_mpo.rs +++ b/src/patched_mpo.rs @@ -217,7 +217,9 @@ impl PatchedMpoPair { let global_left = left_train.clone(); let global_right = right_train.clone(); let tree_input_options = tree::PatchingOptions { - rtol: input_rtol, + // `tree::PatchingOptions` takes a discarded-weight cutoff; `rtol` + // was the same budget as `rtol^2`. + cutoff: input_rtol * input_rtol, max_bond_dim: Some(patch_max_bond), patch_order: left_order.clone(), split_strategy: tree::PatchSplitStrategy::Sequential, @@ -233,7 +235,7 @@ impl PatchedMpoPair { split_strategy: tt::PatchSplitStrategy::Sequential, }; let tree_options = tree::PatchingOptions { - rtol: contract_rtol, + cutoff: contract_rtol * contract_rtol, max_bond_dim: Some(patch_max_bond), patch_order: left_order, split_strategy: tree::PatchSplitStrategy::Sequential, @@ -389,7 +391,7 @@ impl PatchedMpoPair { output_max_bond: usize, ) -> anyhow::Result> { let output_options = tree::PatchingOptions { - rtol, + cutoff: rtol * rtol, max_bond_dim: Some(output_max_bond), ..self.tree_options.clone() }; From 05ebd78bdc40e09f1534a4bab2ac72131ce271d3 Mon Sep 17 00:00:00 2001 From: bench Date: Sun, 11 Oct 2026 07:45:55 +0900 Subject: [PATCH 3/4] Add an optional multithreaded scaling sweep and fix the input cache revision --- AGENTS.md | 8 +++++++- README.md | 8 +++++++- scripts/run_all.sh | 29 +++++++++++++++++++++-------- src/gaussian_input.rs | 2 +- 4 files changed, 36 insertions(+), 11 deletions(-) diff --git a/AGENTS.md b/AGENTS.md index 3ed4e43..c66be6e 100644 --- a/AGENTS.md +++ b/AGENTS.md @@ -14,7 +14,13 @@ ACI residual tolerance is not an L2 tolerance. Keep its internal metric explicit Time only the benchmarked elementwise or contraction operation. Input construction, cache I/O, compression, conversion, patch preparation, output conversion, and accuracy evaluation stay outside timed regions. -Run release builds with one pinned CPU core and single-threaded Rayon and BLAS. Warm up any JIT-compiled external implementation before timing it. +Run release builds with one pinned CPU core and single-threaded Rayon and BLAS. That single-core sweep is the source of truth, and its profile name is the machine alone. Warm up any JIT-compiled external implementation before timing it. + +A machine may carry one additional scaling profile named `-t`, +produced with `BENCH_THREADS=n`, which pins a contiguous core range and lets the +same three cases run on `n` Rayon and BLAS threads. Its `run.yaml` records the +thread count and the core range. Treat it as scaling evidence only: the +single-core profile stays the reference, and the two profiles are never merged. ## Results diff --git a/README.md b/README.md index 5146804..b0721a5 100644 --- a/README.md +++ b/README.md @@ -55,7 +55,13 @@ Requirements are Rust, HDF5, and a BLAS/LAPACK implementation. A complete single BENCH_CPU_CORE=0 scripts/run_all.sh linux-epyc-7713p ``` -The script pins Rayon and common BLAS implementations to one thread, runs the three maintained binaries in release mode, writes raw records under `result//raw/`, records machine metadata in `run.yaml`, and generates `report.md`. +The script pins Rayon and common BLAS implementations and the tenferro thread count to one thread, runs the three maintained binaries in release mode, writes raw records under `result//raw/`, records machine metadata in `run.yaml`, and generates `report.md`. + +Set `BENCH_THREADS=n` for the optional scaling sweep. It uses `n` threads on the contiguous core range starting at `BENCH_CPU_CORE`, and belongs in its own profile named `-t`. The single-core profile stays the source of truth. + +```bash +BENCH_CPU_CORE=0 BENCH_THREADS=4 scripts/run_all.sh linux-epyc-7713p-4t +``` For a probe run: diff --git a/scripts/run_all.sh b/scripts/run_all.sh index 5373181..b78e115 100755 --- a/scripts/run_all.sh +++ b/scripts/run_all.sh @@ -10,19 +10,32 @@ profile=$1 out="result/$profile" raw="$out/raw" core=${BENCH_CPU_CORE:-0} +threads=${BENCH_THREADS:-1} +if ((threads < 1)); then + echo "BENCH_THREADS must be at least 1, got $threads" >&2 + exit 2 +fi +# One pinned core for the source-of-truth sweep, a contiguous range for the +# optional `-t` scaling profile. +if ((threads == 1)); then + affinity="$core" +else + affinity="$core-$((core + threads - 1))" +fi mkdir -p "$raw" rm -f "$raw"/*.json -export RAYON_NUM_THREADS=1 -export OMP_NUM_THREADS=1 -export OPENBLAS_NUM_THREADS=1 -export MKL_NUM_THREADS=1 -export VECLIB_MAXIMUM_THREADS=1 +export RAYON_NUM_THREADS="$threads" +export OMP_NUM_THREADS="$threads" +export OPENBLAS_NUM_THREADS="$threads" +export MKL_NUM_THREADS="$threads" +export VECLIB_MAXIMUM_THREADS="$threads" +export TENFERRO_NUM_THREADS="$threads" export OUT_DIR="$raw" run() { if command -v taskset >/dev/null 2>&1; then - taskset -c "$core" "$@" + taskset -c "$affinity" "$@" else "$@" fi @@ -45,8 +58,8 @@ chip: ${chip:-unknown} memory_gb: ${memory_gb:-unknown} repo_rev: $repo_rev tensor4all_rs_rev: a5157385f58048355a28a6a003d4cb63d3b1349b -threads: 1 -cpu_affinity: $core +threads: $threads +cpu_affinity: $affinity EOF run cargo run --release --locked --bin elementwise_fourier diff --git a/src/gaussian_input.rs b/src/gaussian_input.rs index 04f2d10..fe87606 100644 --- a/src/gaussian_input.rs +++ b/src/gaussian_input.rs @@ -16,7 +16,7 @@ pub const PATCH_CAP: usize = 128; /// Final global relative-L2 input tolerance. pub const INPUT_L2_RTOL: f64 = 1e-6; const CACHE_SCHEMA: &str = "interpolative-gaussian-pair-v2"; -const TENSOR4ALL_REV: &str = "9e9aedaebe0d3918b34dd399ff0981e337f3835b"; +const TENSOR4ALL_REV: &str = "a5157385f58048355a28a6a003d4cb63d3b1349b"; /// Parameters that uniquely define one cached input pair. #[derive(Clone, Debug)] From 5db10f1b688e076a3f00ae77898146d40e6386ae Mon Sep 17 00:00:00 2001 From: bench Date: Sun, 11 Oct 2026 08:09:00 +0900 Subject: [PATCH 4/4] Record the a515738 sweep on the Ryzen AI 9 HX 470, one and four threads --- .../raw/elementwise_fourier-aci-k128.json | 54 ++++++++++++++ .../raw/elementwise_fourier-aci-k16.json | 54 ++++++++++++++ .../raw/elementwise_fourier-aci-k32.json | 54 ++++++++++++++ .../raw/elementwise_fourier-aci-k4.json | 54 ++++++++++++++ .../raw/elementwise_fourier-aci-k64.json | 54 ++++++++++++++ .../raw/elementwise_fourier-aci-k8.json | 54 ++++++++++++++ .../raw/elementwise_fourier-fit-k128.json | 54 ++++++++++++++ .../raw/elementwise_fourier-fit-k16.json | 54 ++++++++++++++ .../raw/elementwise_fourier-fit-k32.json | 54 ++++++++++++++ .../raw/elementwise_fourier-fit-k4.json | 54 ++++++++++++++ .../raw/elementwise_fourier-fit-k64.json | 54 ++++++++++++++ .../raw/elementwise_fourier-fit-k8.json | 54 ++++++++++++++ .../raw/elementwise_fourier-naive-k128.json | 54 ++++++++++++++ .../raw/elementwise_fourier-naive-k16.json | 54 ++++++++++++++ .../raw/elementwise_fourier-naive-k32.json | 54 ++++++++++++++ .../raw/elementwise_fourier-naive-k4.json | 54 ++++++++++++++ .../raw/elementwise_fourier-naive-k64.json | 54 ++++++++++++++ .../raw/elementwise_fourier-naive-k8.json | 54 ++++++++++++++ .../raw/elementwise_fourier-zipup-k128.json | 54 ++++++++++++++ .../raw/elementwise_fourier-zipup-k16.json | 54 ++++++++++++++ .../raw/elementwise_fourier-zipup-k32.json | 54 ++++++++++++++ .../raw/elementwise_fourier-zipup-k4.json | 54 ++++++++++++++ .../raw/elementwise_fourier-zipup-k64.json | 54 ++++++++++++++ .../raw/elementwise_fourier-zipup-k8.json | 54 ++++++++++++++ ...gaussian_elementwise-global_aci-chi16.json | 60 ++++++++++++++++ ...gaussian_elementwise-global_aci-chi24.json | 61 ++++++++++++++++ ...gaussian_elementwise-global_aci-chi41.json | 62 ++++++++++++++++ ...gaussian_elementwise-global_aci-chi64.json | 63 +++++++++++++++++ ...gaussian_elementwise-global_fit-chi16.json | 61 ++++++++++++++++ ...gaussian_elementwise-global_fit-chi24.json | 62 ++++++++++++++++ ...gaussian_elementwise-global_fit-chi41.json | 63 +++++++++++++++++ ...gaussian_elementwise-global_fit-chi64.json | 64 +++++++++++++++++ ...aussian_elementwise-patched_aci-chi16.json | 60 ++++++++++++++++ ...aussian_elementwise-patched_aci-chi24.json | 60 ++++++++++++++++ ...aussian_elementwise-patched_aci-chi41.json | 60 ++++++++++++++++ ...aussian_elementwise-patched_aci-chi64.json | 60 ++++++++++++++++ ...aussian_elementwise-patched_fit-chi16.json | 60 ++++++++++++++++ ...aussian_elementwise-patched_fit-chi24.json | 60 ++++++++++++++++ ...aussian_elementwise-patched_fit-chi41.json | 60 ++++++++++++++++ ...aussian_elementwise-patched_fit-chi64.json | 60 ++++++++++++++++ ...sian_mpo_contraction-global_fit-chi16.json | 64 +++++++++++++++++ ...sian_mpo_contraction-global_fit-chi24.json | 65 +++++++++++++++++ ...sian_mpo_contraction-global_fit-chi41.json | 66 +++++++++++++++++ ...sian_mpo_contraction-global_fit-chi64.json | 67 ++++++++++++++++++ ...ian_mpo_contraction-patched_fit-chi16.json | 63 +++++++++++++++++ ...ian_mpo_contraction-patched_fit-chi24.json | 63 +++++++++++++++++ ...ian_mpo_contraction-patched_fit-chi41.json | 63 +++++++++++++++++ ...ian_mpo_contraction-patched_fit-chi64.json | 63 +++++++++++++++++ result/linux-ryzen-ai-9-hx-470-4t/report.md | 70 +++++++++++++++++++ result/linux-ryzen-ai-9-hx-470-4t/run.yaml | 9 +++ .../raw/elementwise_fourier-aci-k128.json | 54 ++++++++++++++ .../raw/elementwise_fourier-aci-k16.json | 54 ++++++++++++++ .../raw/elementwise_fourier-aci-k32.json | 54 ++++++++++++++ .../raw/elementwise_fourier-aci-k4.json | 54 ++++++++++++++ .../raw/elementwise_fourier-aci-k64.json | 54 ++++++++++++++ .../raw/elementwise_fourier-aci-k8.json | 54 ++++++++++++++ .../raw/elementwise_fourier-fit-k128.json | 54 ++++++++++++++ .../raw/elementwise_fourier-fit-k16.json | 54 ++++++++++++++ .../raw/elementwise_fourier-fit-k32.json | 54 ++++++++++++++ .../raw/elementwise_fourier-fit-k4.json | 54 ++++++++++++++ .../raw/elementwise_fourier-fit-k64.json | 54 ++++++++++++++ .../raw/elementwise_fourier-fit-k8.json | 54 ++++++++++++++ .../raw/elementwise_fourier-naive-k128.json | 54 ++++++++++++++ .../raw/elementwise_fourier-naive-k16.json | 54 ++++++++++++++ .../raw/elementwise_fourier-naive-k32.json | 54 ++++++++++++++ .../raw/elementwise_fourier-naive-k4.json | 54 ++++++++++++++ .../raw/elementwise_fourier-naive-k64.json | 54 ++++++++++++++ .../raw/elementwise_fourier-naive-k8.json | 54 ++++++++++++++ .../raw/elementwise_fourier-zipup-k128.json | 54 ++++++++++++++ .../raw/elementwise_fourier-zipup-k16.json | 54 ++++++++++++++ .../raw/elementwise_fourier-zipup-k32.json | 54 ++++++++++++++ .../raw/elementwise_fourier-zipup-k4.json | 54 ++++++++++++++ .../raw/elementwise_fourier-zipup-k64.json | 54 ++++++++++++++ .../raw/elementwise_fourier-zipup-k8.json | 54 ++++++++++++++ ...gaussian_elementwise-global_aci-chi16.json | 60 ++++++++++++++++ ...gaussian_elementwise-global_aci-chi24.json | 61 ++++++++++++++++ ...gaussian_elementwise-global_aci-chi41.json | 62 ++++++++++++++++ ...gaussian_elementwise-global_aci-chi64.json | 63 +++++++++++++++++ ...gaussian_elementwise-global_fit-chi16.json | 61 ++++++++++++++++ ...gaussian_elementwise-global_fit-chi24.json | 62 ++++++++++++++++ ...gaussian_elementwise-global_fit-chi41.json | 63 +++++++++++++++++ ...gaussian_elementwise-global_fit-chi64.json | 64 +++++++++++++++++ ...aussian_elementwise-patched_aci-chi16.json | 60 ++++++++++++++++ ...aussian_elementwise-patched_aci-chi24.json | 60 ++++++++++++++++ ...aussian_elementwise-patched_aci-chi41.json | 60 ++++++++++++++++ ...aussian_elementwise-patched_aci-chi64.json | 60 ++++++++++++++++ ...aussian_elementwise-patched_fit-chi16.json | 60 ++++++++++++++++ ...aussian_elementwise-patched_fit-chi24.json | 60 ++++++++++++++++ ...aussian_elementwise-patched_fit-chi41.json | 60 ++++++++++++++++ ...aussian_elementwise-patched_fit-chi64.json | 60 ++++++++++++++++ ...sian_mpo_contraction-global_fit-chi16.json | 64 +++++++++++++++++ ...sian_mpo_contraction-global_fit-chi24.json | 65 +++++++++++++++++ ...sian_mpo_contraction-global_fit-chi41.json | 66 +++++++++++++++++ ...sian_mpo_contraction-global_fit-chi64.json | 67 ++++++++++++++++++ ...ian_mpo_contraction-patched_fit-chi16.json | 63 +++++++++++++++++ ...ian_mpo_contraction-patched_fit-chi24.json | 63 +++++++++++++++++ ...ian_mpo_contraction-patched_fit-chi41.json | 63 +++++++++++++++++ ...ian_mpo_contraction-patched_fit-chi64.json | 63 +++++++++++++++++ result/linux-ryzen-ai-9-hx-470/report.md | 70 +++++++++++++++++++ result/linux-ryzen-ai-9-hx-470/run.yaml | 9 +++ 100 files changed, 5730 insertions(+) create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k128.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k16.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k32.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k4.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k64.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k8.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k128.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k16.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k32.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k4.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k64.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k8.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k128.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k16.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k32.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k4.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k64.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k8.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k128.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k16.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k32.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k4.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k64.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k8.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_aci-chi16.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_aci-chi24.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_aci-chi41.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_aci-chi64.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_fit-chi16.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_fit-chi24.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_fit-chi41.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_fit-chi64.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_aci-chi16.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_aci-chi24.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_aci-chi41.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_aci-chi64.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_fit-chi16.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_fit-chi24.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_fit-chi41.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_fit-chi64.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-global_fit-chi16.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-global_fit-chi24.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-global_fit-chi41.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-global_fit-chi64.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-patched_fit-chi16.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-patched_fit-chi24.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-patched_fit-chi41.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-patched_fit-chi64.json create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/report.md create mode 100644 result/linux-ryzen-ai-9-hx-470-4t/run.yaml create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k128.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k16.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k32.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k4.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k64.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k8.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k128.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k16.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k32.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k4.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k64.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k8.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k128.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k16.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k32.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k4.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k64.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k8.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k128.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k16.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k32.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k4.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k64.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k8.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_aci-chi16.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_aci-chi24.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_aci-chi41.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_aci-chi64.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_fit-chi16.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_fit-chi24.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_fit-chi41.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_fit-chi64.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_aci-chi16.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_aci-chi24.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_aci-chi41.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_aci-chi64.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_fit-chi16.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_fit-chi24.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_fit-chi41.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_fit-chi64.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-global_fit-chi16.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-global_fit-chi24.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-global_fit-chi41.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-global_fit-chi64.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-patched_fit-chi16.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-patched_fit-chi24.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-patched_fit-chi41.json create mode 100644 result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-patched_fit-chi64.json create mode 100644 result/linux-ryzen-ai-9-hx-470/report.md create mode 100644 result/linux-ryzen-ai-9-hx-470/run.yaml diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k128.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k128.json new file mode 100644 index 0000000..a892be7 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k128.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "aci", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "aci_absolute_residual", + "k_max": 128, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.075586995, + "wall_times_secs": [ + 0.077039697, + 0.075586995, + 0.075348304, + 0.075582406, + 0.075701737 + ], + "max_error": 3.246891024421058e-9, + "input_max_bond_dim": 16, + "output_max_bond_dim": 32, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 32, + 30, + 24, + 17, + 14, + 10, + 8, + 6, + 6, + 4, + 4, + 4, + 4, + 3, + 2 + ], + "n_params": 6808, + "n_allocations": 1115933, + "allocated_bytes": 142524978 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k16.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k16.json new file mode 100644 index 0000000..186bfff --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k16.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "aci", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "aci_absolute_residual", + "k_max": 16, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.066019249, + "wall_times_secs": [ + 0.066010253, + 0.067843437, + 0.066019249, + 0.065421827, + 0.0674392 + ], + "max_error": 6.085556560690714e-9, + "input_max_bond_dim": 8, + "output_max_bond_dim": 16, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 16, + 13, + 9, + 9, + 7, + 7, + 5, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 2 + ], + "n_params": 2206, + "n_allocations": 1095666, + "allocated_bytes": 116687466 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k32.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k32.json new file mode 100644 index 0000000..a3c9d92 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k32.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "aci", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "aci_absolute_residual", + "k_max": 32, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.077826996, + "wall_times_secs": [ + 0.077699922, + 0.077826996, + 0.077555637, + 0.07879775, + 0.082593808 + ], + "max_error": 4.930920676253506e-9, + "input_max_bond_dim": 11, + "output_max_bond_dim": 23, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 23, + 17, + 14, + 10, + 9, + 7, + 7, + 5, + 5, + 5, + 4, + 4, + 3, + 3, + 2 + ], + "n_params": 3318, + "n_allocations": 1253380, + "allocated_bytes": 142825089 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k4.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k4.json new file mode 100644 index 0000000..2eb6cae --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k4.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "aci", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "aci_absolute_residual", + "k_max": 4, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.056474199, + "wall_times_secs": [ + 0.056590512, + 0.056474199, + 0.055539893, + 0.056709552, + 0.054970011 + ], + "max_error": 1.1659684028740639e-8, + "input_max_bond_dim": 5, + "output_max_bond_dim": 9, + "output_bond_dims": [ + 2, + 4, + 8, + 9, + 8, + 8, + 7, + 7, + 5, + 5, + 4, + 4, + 4, + 3, + 3, + 3, + 2, + 2, + 2 + ], + "n_params": 1026, + "n_allocations": 935094, + "allocated_bytes": 96798050 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k64.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k64.json new file mode 100644 index 0000000..a7ba327 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k64.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "aci", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "aci_absolute_residual", + "k_max": 64, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.074521806, + "wall_times_secs": [ + 0.074521806, + 0.073874522, + 0.074008339, + 0.078656239, + 0.079269852 + ], + "max_error": 4.479701266618146e-9, + "input_max_bond_dim": 14, + "output_max_bond_dim": 29, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 29, + 22, + 17, + 14, + 10, + 9, + 7, + 7, + 5, + 5, + 5, + 5, + 4, + 3, + 2 + ], + "n_params": 4752, + "n_allocations": 1149587, + "allocated_bytes": 136142407 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k8.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k8.json new file mode 100644 index 0000000..f0696e1 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-aci-k8.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "aci", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "aci_absolute_residual", + "k_max": 8, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.077047094, + "wall_times_secs": [ + 0.080023649, + 0.077047094, + 0.076721664, + 0.076725641, + 0.07835537 + ], + "max_error": 8.181184498521454e-9, + "input_max_bond_dim": 8, + "output_max_bond_dim": 16, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 13, + 9, + 9, + 7, + 7, + 5, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 2, + 2 + ], + "n_params": 1702, + "n_allocations": 1293260, + "allocated_bytes": 135781184 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k128.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k128.json new file mode 100644 index 0000000..c58814e --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k128.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "fit", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 128, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.023957656, + "wall_times_secs": [ + 0.027386345, + 0.023897765, + 0.024183934, + 0.023957656, + 0.023171756 + ], + "max_error": 1.0440436475655285e-8, + "input_max_bond_dim": 16, + "output_max_bond_dim": 18, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 18, + 13, + 10, + 8, + 7, + 6, + 5, + 4, + 4, + 4, + 3, + 3, + 3, + 3, + 2 + ], + "n_params": 2258, + "n_allocations": 145567, + "allocated_bytes": 33538018 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k16.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k16.json new file mode 100644 index 0000000..1ab9b0d --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k16.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "fit", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 16, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.019670291, + "wall_times_secs": [ + 0.019670291, + 0.020781742, + 0.019494146, + 0.019081725, + 0.02061783 + ], + "max_error": 1.6203459388257622e-8, + "input_max_bond_dim": 8, + "output_max_bond_dim": 10, + "output_bond_dims": [ + 2, + 4, + 8, + 10, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2 + ], + "n_params": 900, + "n_allocations": 143319, + "allocated_bytes": 26154736 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k32.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k32.json new file mode 100644 index 0000000..9152544 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k32.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "fit", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 32, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.019987746, + "wall_times_secs": [ + 0.020007463, + 0.019959866, + 0.020464336, + 0.019987746, + 0.019660182 + ], + "max_error": 1.8064612929906997e-8, + "input_max_bond_dim": 11, + "output_max_bond_dim": 13, + "output_bond_dims": [ + 2, + 4, + 8, + 13, + 10, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2, + 2 + ], + "n_params": 1200, + "n_allocations": 144030, + "allocated_bytes": 27614348 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k4.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k4.json new file mode 100644 index 0000000..49a7dee --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k4.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "fit", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 4, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.025641603, + "wall_times_secs": [ + 0.026391807, + 0.025641603, + 0.02837619, + 0.0226228, + 0.019427151 + ], + "max_error": 1.213100714766008e-8, + "input_max_bond_dim": 5, + "output_max_bond_dim": 8, + "output_bond_dims": [ + 2, + 4, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2, + 2 + ], + "n_params": 606, + "n_allocations": 142954, + "allocated_bytes": 24796024 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k64.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k64.json new file mode 100644 index 0000000..3b82179 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k64.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "fit", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 64, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.021687495, + "wall_times_secs": [ + 0.023494721, + 0.020561306, + 0.020997812, + 0.021687495, + 0.022911284 + ], + "max_error": 1.8126117360890213e-8, + "input_max_bond_dim": 14, + "output_max_bond_dim": 16, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 13, + 10, + 8, + 7, + 6, + 5, + 4, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2 + ], + "n_params": 1638, + "n_allocations": 143911, + "allocated_bytes": 29442952 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k8.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k8.json new file mode 100644 index 0000000..f7875b1 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-fit-k8.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "fit", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 8, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.018698436, + "wall_times_secs": [ + 0.018845999, + 0.018698436, + 0.018325108, + 0.019347133, + 0.018482608 + ], + "max_error": 1.0663293755762709e-8, + "input_max_bond_dim": 8, + "output_max_bond_dim": 8, + "output_bond_dims": [ + 2, + 4, + 8, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2 + ], + "n_params": 726, + "n_allocations": 143245, + "allocated_bytes": 25437613 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k128.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k128.json new file mode 100644 index 0000000..b693e56 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k128.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "naive", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 128, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.009561731, + "wall_times_secs": [ + 0.009595293, + 0.009561731, + 0.009437492, + 0.009665663, + 0.009354137 + ], + "max_error": 1.0441367943793766e-8, + "input_max_bond_dim": 16, + "output_max_bond_dim": 18, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 18, + 13, + 10, + 8, + 7, + 6, + 5, + 4, + 4, + 4, + 3, + 3, + 3, + 3, + 2 + ], + "n_params": 2258, + "n_allocations": 7403, + "allocated_bytes": 24647767 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k16.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k16.json new file mode 100644 index 0000000..4465071 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k16.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "naive", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 16, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.002000402, + "wall_times_secs": [ + 0.002000402, + 0.001970176, + 0.001946714, + 0.002104856, + 0.002033303 + ], + "max_error": 1.62021859796595e-8, + "input_max_bond_dim": 8, + "output_max_bond_dim": 10, + "output_bond_dims": [ + 2, + 4, + 8, + 10, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2 + ], + "n_params": 900, + "n_allocations": 5644, + "allocated_bytes": 5376755 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k32.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k32.json new file mode 100644 index 0000000..570e95f --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k32.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "naive", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 32, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.003919836, + "wall_times_secs": [ + 0.003309209, + 0.003461259, + 0.003919836, + 0.006673749, + 0.007727866 + ], + "max_error": 1.806346087785027e-8, + "input_max_bond_dim": 11, + "output_max_bond_dim": 13, + "output_bond_dims": [ + 2, + 4, + 8, + 13, + 10, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2, + 2 + ], + "n_params": 1200, + "n_allocations": 6139, + "allocated_bytes": 8621382 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k4.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k4.json new file mode 100644 index 0000000..6d78d22 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k4.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "naive", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 4, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.0015822, + "wall_times_secs": [ + 0.001715927, + 0.0015822, + 0.001556814, + 0.001560801, + 0.001604251 + ], + "max_error": 1.2110321672966727e-8, + "input_max_bond_dim": 5, + "output_max_bond_dim": 8, + "output_bond_dims": [ + 2, + 4, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2, + 2 + ], + "n_params": 606, + "n_allocations": 5467, + "allocated_bytes": 2737219 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k64.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k64.json new file mode 100644 index 0000000..42a0c63 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k64.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "naive", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 64, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.00541691, + "wall_times_secs": [ + 0.005155397, + 0.005436696, + 0.005590741, + 0.005330049, + 0.00541691 + ], + "max_error": 1.8125685207254637e-8, + "input_max_bond_dim": 14, + "output_max_bond_dim": 16, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 13, + 10, + 8, + 7, + 6, + 5, + 4, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2 + ], + "n_params": 1638, + "n_allocations": 6685, + "allocated_bytes": 14418468 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k8.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k8.json new file mode 100644 index 0000000..a4b54bf --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-naive-k8.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "naive", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 8, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.001533, + "wall_times_secs": [ + 0.001542408, + 0.001547887, + 0.001530445, + 0.001533, + 0.001483058 + ], + "max_error": 1.0657100928094615e-8, + "input_max_bond_dim": 8, + "output_max_bond_dim": 8, + "output_bond_dims": [ + 2, + 4, + 8, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2 + ], + "n_params": 726, + "n_allocations": 5599, + "allocated_bytes": 4039823 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k128.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k128.json new file mode 100644 index 0000000..0c9e926 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k128.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "zipup", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 128, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.016043106, + "wall_times_secs": [ + 0.013146509, + 0.016043106, + 0.013183156, + 0.018653615, + 0.016457811 + ], + "max_error": 1.0638874285729503e-8, + "input_max_bond_dim": 16, + "output_max_bond_dim": 18, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 18, + 13, + 10, + 8, + 7, + 6, + 5, + 4, + 4, + 4, + 3, + 3, + 3, + 3, + 2 + ], + "n_params": 2258, + "n_allocations": 83392, + "allocated_bytes": 19139558 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k16.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k16.json new file mode 100644 index 0000000..872752c --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k16.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "zipup", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 16, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.010695102, + "wall_times_secs": [ + 0.010510161, + 0.010327784, + 0.010922833, + 0.010695102, + 0.010697567 + ], + "max_error": 1.636334768856949e-8, + "input_max_bond_dim": 8, + "output_max_bond_dim": 10, + "output_bond_dims": [ + 2, + 4, + 8, + 10, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2 + ], + "n_params": 900, + "n_allocations": 80869, + "allocated_bytes": 13720830 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k32.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k32.json new file mode 100644 index 0000000..87ae243 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k32.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "zipup", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 32, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.011347687, + "wall_times_secs": [ + 0.011370639, + 0.011406486, + 0.011259474, + 0.011347687, + 0.011306742 + ], + "max_error": 2.0007564089687593e-8, + "input_max_bond_dim": 11, + "output_max_bond_dim": 13, + "output_bond_dims": [ + 2, + 4, + 8, + 13, + 10, + 8, + 7, + 6, + 5, + 4, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2, + 2 + ], + "n_params": 1182, + "n_allocations": 81544, + "allocated_bytes": 14732474 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k4.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k4.json new file mode 100644 index 0000000..b0ec7b1 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k4.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "zipup", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 4, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.014576797, + "wall_times_secs": [ + 0.014800992, + 0.015854656, + 0.014342325, + 0.014576797, + 0.014096541 + ], + "max_error": 1.1974477610213059e-8, + "input_max_bond_dim": 5, + "output_max_bond_dim": 8, + "output_bond_dims": [ + 2, + 4, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2, + 2 + ], + "n_params": 606, + "n_allocations": 80520, + "allocated_bytes": 12695893 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k64.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k64.json new file mode 100644 index 0000000..b131066 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k64.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "zipup", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 64, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.011935392, + "wall_times_secs": [ + 0.01252984, + 0.011935392, + 0.013106886, + 0.011520166, + 0.011660465 + ], + "max_error": 1.8557198428833864e-8, + "input_max_bond_dim": 14, + "output_max_bond_dim": 16, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 13, + 10, + 8, + 7, + 6, + 5, + 4, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2 + ], + "n_params": 1638, + "n_allocations": 81873, + "allocated_bytes": 16025684 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k8.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k8.json new file mode 100644 index 0000000..e84c00a --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/elementwise_fourier-zipup-k8.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "zipup", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 8, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.009921083, + "wall_times_secs": [ + 0.010263806, + 0.009921083, + 0.009907048, + 0.009912257, + 0.010021369 + ], + "max_error": 1.0797260594548758e-8, + "input_max_bond_dim": 8, + "output_max_bond_dim": 8, + "output_bond_dims": [ + 2, + 4, + 8, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2 + ], + "n_params": 726, + "n_allocations": 80806, + "allocated_bytes": 13190835 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_aci-chi16.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_aci-chi16.json new file mode 100644 index 0000000..fa85476 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_aci-chi16.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "global_aci", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.10606601717798216, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n2-r6-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004632221, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.005599567, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 16, + "input_params": 3740, + "input_patch_count": 2, + "input_patch_params": 3740, + "internal_tolerance": 1e-8, + "internal_tolerance_metric": "aci_scale_relative_residual", + "interpolation_tolerance": 1e-10, + "left_chi": 16, + "left_input_sampled_relative_l2": 1.2777851899668411e-6, + "n_gauss": 2, + "patch_build_secs": 0.011786087, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 6, + "raw_left_chi": 32, + "raw_left_params": 4640, + "raw_right_chi": 26, + "raw_right_params": 3872, + "rho_max": 8.0, + "right_chi": 16, + "right_input_sampled_relative_l2": 9.428700715139774e-7, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.004898823, + "wall_times_secs": [ + 0.004898823 + ], + "max_error": 1.3982004556389037e-6, + "input_max_bond_dim": 16, + "output_max_bond_dim": 47, + "output_bond_dims": [ + 4, + 16, + 47, + 16, + 4 + ], + "n_params": 6560, + "input_build_secs": 0.0, + "n_allocations": 57207, + "allocated_bytes": 12091644 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_aci-chi24.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_aci-chi24.json new file mode 100644 index 0000000..00a08f9 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_aci-chi24.json @@ -0,0 +1,61 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "global_aci", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.2121320343559643, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n8-r7-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004221493, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.002755687, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 24, + "input_params": 7924, + "input_patch_count": 2, + "input_patch_params": 7924, + "internal_tolerance": 1e-8, + "internal_tolerance_metric": "aci_scale_relative_residual", + "interpolation_tolerance": 1e-10, + "left_chi": 23, + "left_input_sampled_relative_l2": 1.7758182496310817e-6, + "n_gauss": 8, + "patch_build_secs": 0.015475047, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 7, + "raw_left_chi": 50, + "raw_left_params": 12720, + "raw_right_chi": 50, + "raw_right_params": 12720, + "rho_max": 8.0, + "right_chi": 24, + "right_input_sampled_relative_l2": 1.4847847441262137e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.012319292, + "wall_times_secs": [ + 0.012319292 + ], + "max_error": 2.941790247407175e-6, + "input_max_bond_dim": 24, + "output_max_bond_dim": 64, + "output_bond_dims": [ + 4, + 16, + 64, + 59, + 15, + 4 + ], + "n_params": 23268, + "input_build_secs": 0.0, + "n_allocations": 87494, + "allocated_bytes": 30761484 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_aci-chi41.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_aci-chi41.json new file mode 100644 index 0000000..e961a78 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_aci-chi41.json @@ -0,0 +1,62 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "global_aci", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.4242640687119286, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n32-r8-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004694927, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.006029541, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 41, + "input_params": 18380, + "input_patch_count": 2, + "input_patch_params": 18380, + "internal_tolerance": 1e-8, + "internal_tolerance_metric": "aci_scale_relative_residual", + "interpolation_tolerance": 1e-10, + "left_chi": 41, + "left_input_sampled_relative_l2": 1.8640163986326415e-6, + "n_gauss": 32, + "patch_build_secs": 0.022816971, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 8, + "raw_left_chi": 64, + "raw_left_params": 27592, + "raw_right_chi": 64, + "raw_right_params": 27592, + "rho_max": 8.0, + "right_chi": 40, + "right_input_sampled_relative_l2": 2.232276096287149e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.038987893, + "wall_times_secs": [ + 0.038987893 + ], + "max_error": 3.468069058709741e-6, + "input_max_bond_dim": 41, + "output_max_bond_dim": 183, + "output_bond_dims": [ + 4, + 16, + 64, + 183, + 60, + 15, + 4 + ], + "n_params": 98992, + "input_build_secs": 0.0, + "n_allocations": 110020, + "allocated_bytes": 82716230 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_aci-chi64.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_aci-chi64.json new file mode 100644 index 0000000..c66c64e --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_aci-chi64.json @@ -0,0 +1,63 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "global_aci", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.8485281374238572, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n128-r9-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.005456924, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.022905113, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 64, + "input_params": 43200, + "input_patch_count": 2, + "input_patch_params": 43200, + "internal_tolerance": 1e-8, + "internal_tolerance_metric": "aci_scale_relative_residual", + "interpolation_tolerance": 1e-10, + "left_chi": 64, + "left_input_sampled_relative_l2": 2.1682089914631927e-6, + "n_gauss": 128, + "patch_build_secs": 0.037030641, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 9, + "raw_left_chi": 99, + "raw_left_params": 62240, + "raw_right_chi": 98, + "raw_right_params": 60416, + "rho_max": 8.0, + "right_chi": 64, + "right_input_sampled_relative_l2": 2.8526222942130557e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.148519914, + "wall_times_secs": [ + 0.148519914 + ], + "max_error": 2.7045739440128625e-6, + "input_max_bond_dim": 64, + "output_max_bond_dim": 256, + "output_bond_dims": [ + 4, + 16, + 64, + 256, + 193, + 60, + 15, + 4 + ], + "n_params": 317712, + "input_build_secs": 0.0, + "n_allocations": 141977, + "allocated_bytes": 541306696 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_fit-chi16.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_fit-chi16.json new file mode 100644 index 0000000..bf199bf --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_fit-chi16.json @@ -0,0 +1,61 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "global_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.10606601717798216, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n2-r6-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004632221, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.005599567, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 16, + "input_params": 3740, + "input_patch_count": 2, + "input_patch_params": 3740, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 16, + "left_input_sampled_relative_l2": 1.2777851899668411e-6, + "n_gauss": 2, + "patch_build_secs": 0.011786087, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 6, + "raw_left_chi": 32, + "raw_left_params": 4640, + "raw_right_chi": 26, + "raw_right_params": 3872, + "rho_max": 8.0, + "right_chi": 16, + "right_input_sampled_relative_l2": 9.428700715139774e-7, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.010786432, + "wall_times_secs": [ + 0.010786432 + ], + "max_error": 2.2930348001465887e-6, + "input_max_bond_dim": 16, + "output_max_bond_dim": 17, + "output_bond_dims": [ + 4, + 16, + 17, + 11, + 4 + ], + "n_params": 2300, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 60186, + "allocated_bytes": 120470829 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_fit-chi24.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_fit-chi24.json new file mode 100644 index 0000000..18ed1f0 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_fit-chi24.json @@ -0,0 +1,62 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "global_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.2121320343559643, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n8-r7-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004221493, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.002755687, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 24, + "input_params": 7924, + "input_patch_count": 2, + "input_patch_params": 7924, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 23, + "left_input_sampled_relative_l2": 1.7758182496310817e-6, + "n_gauss": 8, + "patch_build_secs": 0.015475047, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 7, + "raw_left_chi": 50, + "raw_left_params": 12720, + "raw_right_chi": 50, + "raw_right_params": 12720, + "rho_max": 8.0, + "right_chi": 24, + "right_input_sampled_relative_l2": 1.4847847441262137e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.017571738, + "wall_times_secs": [ + 0.017571738 + ], + "max_error": 3.4615301397236164e-6, + "input_max_bond_dim": 24, + "output_max_bond_dim": 27, + "output_bond_dims": [ + 4, + 16, + 27, + 17, + 11, + 4 + ], + "n_params": 4776, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 70531, + "allocated_bytes": 27815926 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_fit-chi41.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_fit-chi41.json new file mode 100644 index 0000000..f8c97bd --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_fit-chi41.json @@ -0,0 +1,63 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "global_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.4242640687119286, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n32-r8-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004694927, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.006029541, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 41, + "input_params": 18380, + "input_patch_count": 2, + "input_patch_params": 18380, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 41, + "left_input_sampled_relative_l2": 1.8640163986326415e-6, + "n_gauss": 32, + "patch_build_secs": 0.022816971, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 8, + "raw_left_chi": 64, + "raw_left_params": 27592, + "raw_right_chi": 64, + "raw_right_params": 27592, + "rho_max": 8.0, + "right_chi": 40, + "right_input_sampled_relative_l2": 2.232276096287149e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.042459182, + "wall_times_secs": [ + 0.042459182 + ], + "max_error": 4.661153609314938e-6, + "input_max_bond_dim": 41, + "output_max_bond_dim": 48, + "output_bond_dims": [ + 4, + 16, + 48, + 27, + 16, + 10, + 4 + ], + "n_params": 11072, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 110564, + "allocated_bytes": 76711637 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_fit-chi64.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_fit-chi64.json new file mode 100644 index 0000000..ddd737a --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-global_fit-chi64.json @@ -0,0 +1,64 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "global_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.8485281374238572, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n128-r9-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.005456924, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.022905113, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 64, + "input_params": 43200, + "input_patch_count": 2, + "input_patch_params": 43200, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 64, + "left_input_sampled_relative_l2": 2.1682089914631927e-6, + "n_gauss": 128, + "patch_build_secs": 0.037030641, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 9, + "raw_left_chi": 99, + "raw_left_params": 62240, + "raw_right_chi": 98, + "raw_right_params": 60416, + "rho_max": 8.0, + "right_chi": 64, + "right_input_sampled_relative_l2": 2.8526222942130557e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.098041509, + "wall_times_secs": [ + 0.098041509 + ], + "max_error": 3.993268189960527e-6, + "input_max_bond_dim": 64, + "output_max_bond_dim": 64, + "output_bond_dims": [ + 4, + 16, + 64, + 51, + 27, + 16, + 10, + 4 + ], + "n_params": 25476, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 140014, + "allocated_bytes": 183575631 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_aci-chi16.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_aci-chi16.json new file mode 100644 index 0000000..2126246 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_aci-chi16.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "patched_aci", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.10606601717798216, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n2-r6-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004632221, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.005599567, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 16, + "input_params": 3740, + "input_patch_count": 2, + "input_patch_params": 3740, + "internal_tolerance": 1e-8, + "internal_tolerance_metric": "aci_absolute_patch_residual_then_global_l2", + "interpolation_tolerance": 1e-10, + "left_chi": 16, + "left_input_sampled_relative_l2": 1.2777851899668411e-6, + "n_gauss": 2, + "patch_build_secs": 0.011786087, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 6, + "raw_left_chi": 32, + "raw_left_params": 4640, + "raw_right_chi": 26, + "raw_right_params": 3872, + "rho_max": 8.0, + "right_chi": 16, + "right_input_sampled_relative_l2": 9.428700715139774e-7, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 0.5771667006962656 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.008487709, + "wall_times_secs": [ + 0.008487709 + ], + "max_error": 2.2917499150632246e-6, + "input_max_bond_dim": 16, + "output_max_bond_dim": 17, + "output_bond_dims": [ + 17 + ], + "n_params": 2300, + "n_patches": 1, + "max_patch_bond": 17, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 81260, + "allocated_bytes": 18097700 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_aci-chi24.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_aci-chi24.json new file mode 100644 index 0000000..5aa9402 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_aci-chi24.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "patched_aci", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.2121320343559643, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n8-r7-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004221493, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.002755687, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 24, + "input_params": 7924, + "input_patch_count": 2, + "input_patch_params": 7924, + "internal_tolerance": 1e-8, + "internal_tolerance_metric": "aci_absolute_patch_residual_then_global_l2", + "interpolation_tolerance": 1e-10, + "left_chi": 23, + "left_input_sampled_relative_l2": 1.7758182496310817e-6, + "n_gauss": 8, + "patch_build_secs": 0.015475047, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 7, + "raw_left_chi": 50, + "raw_left_params": 12720, + "raw_right_chi": 50, + "raw_right_params": 12720, + "rho_max": 8.0, + "right_chi": 24, + "right_input_sampled_relative_l2": 1.4847847441262137e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 0.729933669390724 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.016877276, + "wall_times_secs": [ + 0.016877276 + ], + "max_error": 3.4589838840323777e-6, + "input_max_bond_dim": 24, + "output_max_bond_dim": 28, + "output_bond_dims": [ + 28 + ], + "n_params": 4908, + "n_patches": 1, + "max_patch_bond": 28, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 116482, + "allocated_bytes": 40758559 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_aci-chi41.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_aci-chi41.json new file mode 100644 index 0000000..eea72b5 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_aci-chi41.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "patched_aci", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.4242640687119286, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n32-r8-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004694927, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.006029541, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 41, + "input_params": 18380, + "input_patch_count": 2, + "input_patch_params": 18380, + "internal_tolerance": 1e-8, + "internal_tolerance_metric": "aci_absolute_patch_residual_then_global_l2", + "interpolation_tolerance": 1e-10, + "left_chi": 41, + "left_input_sampled_relative_l2": 1.8640163986326415e-6, + "n_gauss": 32, + "patch_build_secs": 0.022816971, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 8, + "raw_left_chi": 64, + "raw_left_params": 27592, + "raw_right_chi": 64, + "raw_right_params": 27592, + "rho_max": 8.0, + "right_chi": 40, + "right_input_sampled_relative_l2": 2.232276096287149e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 1.2081059645825172 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.032271915, + "wall_times_secs": [ + 0.032271915 + ], + "max_error": 4.365879675737119e-6, + "input_max_bond_dim": 41, + "output_max_bond_dim": 48, + "output_bond_dims": [ + 48 + ], + "n_params": 11740, + "n_patches": 1, + "max_patch_bond": 48, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 117649, + "allocated_bytes": 71827805 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_aci-chi64.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_aci-chi64.json new file mode 100644 index 0000000..f7533e3 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_aci-chi64.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "patched_aci", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.8485281374238572, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n128-r9-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.005456924, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.022905113, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 64, + "input_params": 43200, + "input_patch_count": 2, + "input_patch_params": 43200, + "internal_tolerance": 1e-8, + "internal_tolerance_metric": "aci_absolute_patch_residual_then_global_l2", + "interpolation_tolerance": 1e-10, + "left_chi": 64, + "left_input_sampled_relative_l2": 2.1682089914631927e-6, + "n_gauss": 128, + "patch_build_secs": 0.037030641, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 9, + "raw_left_chi": 99, + "raw_left_params": 62240, + "raw_right_chi": 98, + "raw_right_params": 60416, + "rho_max": 8.0, + "right_chi": 64, + "right_input_sampled_relative_l2": 2.8526222942130557e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 2.033731090120617 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.073028295, + "wall_times_secs": [ + 0.073028295 + ], + "max_error": 3.1872280234314077e-6, + "input_max_bond_dim": 64, + "output_max_bond_dim": 64, + "output_bond_dims": [ + 64 + ], + "n_params": 28028, + "n_patches": 1, + "max_patch_bond": 64, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 150487, + "allocated_bytes": 141534333 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_fit-chi16.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_fit-chi16.json new file mode 100644 index 0000000..3151c6f --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_fit-chi16.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "patched_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.10606601717798216, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n2-r6-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004632221, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.005599567, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 16, + "input_params": 3740, + "input_patch_count": 2, + "input_patch_params": 3740, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 16, + "left_input_sampled_relative_l2": 1.2777851899668411e-6, + "n_gauss": 2, + "patch_build_secs": 0.011786087, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 6, + "raw_left_chi": 32, + "raw_left_params": 4640, + "raw_right_chi": 26, + "raw_right_params": 3872, + "rho_max": 8.0, + "right_chi": 16, + "right_input_sampled_relative_l2": 9.428700715139774e-7, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 0.6216124243279424 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.017352343, + "wall_times_secs": [ + 0.017352343 + ], + "max_error": 2.2917390948877796e-6, + "input_max_bond_dim": 16, + "output_max_bond_dim": 17, + "output_bond_dims": [ + 17 + ], + "n_params": 2300, + "n_patches": 1, + "max_patch_bond": 17, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 84856, + "allocated_bytes": 27330768 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_fit-chi24.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_fit-chi24.json new file mode 100644 index 0000000..80e6efe --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_fit-chi24.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "patched_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.2121320343559643, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n8-r7-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004221493, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.002755687, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 24, + "input_params": 7924, + "input_patch_count": 2, + "input_patch_params": 7924, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 23, + "left_input_sampled_relative_l2": 1.7758182496310817e-6, + "n_gauss": 8, + "patch_build_secs": 0.015475047, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 7, + "raw_left_chi": 50, + "raw_left_params": 12720, + "raw_right_chi": 50, + "raw_right_params": 12720, + "rho_max": 8.0, + "right_chi": 24, + "right_input_sampled_relative_l2": 1.4847847441262137e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 0.4876359601749887 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.036034541, + "wall_times_secs": [ + 0.036034541 + ], + "max_error": 3.458972024601682e-6, + "input_max_bond_dim": 24, + "output_max_bond_dim": 28, + "output_bond_dims": [ + 28 + ], + "n_params": 4908, + "n_patches": 1, + "max_patch_bond": 28, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 123259, + "allocated_bytes": 67324506 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_fit-chi41.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_fit-chi41.json new file mode 100644 index 0000000..446f5e1 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_fit-chi41.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "patched_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.4242640687119286, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n32-r8-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004694927, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.006029541, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 41, + "input_params": 18380, + "input_patch_count": 2, + "input_patch_params": 18380, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 41, + "left_input_sampled_relative_l2": 1.8640163986326415e-6, + "n_gauss": 32, + "patch_build_secs": 0.022816971, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 8, + "raw_left_chi": 64, + "raw_left_params": 27592, + "raw_right_chi": 64, + "raw_right_params": 27592, + "rho_max": 8.0, + "right_chi": 40, + "right_input_sampled_relative_l2": 2.232276096287149e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 0.49829809028917543 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.085208398, + "wall_times_secs": [ + 0.085208398 + ], + "max_error": 4.36543850623382e-6, + "input_max_bond_dim": 41, + "output_max_bond_dim": 48, + "output_bond_dims": [ + 48 + ], + "n_params": 11740, + "n_patches": 1, + "max_patch_bond": 48, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 183635, + "allocated_bytes": 158474682 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_fit-chi64.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_fit-chi64.json new file mode 100644 index 0000000..c42c590 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_elementwise-patched_fit-chi64.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "patched_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.8485281374238572, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n128-r9-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.005456924, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.022905113, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 64, + "input_params": 43200, + "input_patch_count": 2, + "input_patch_params": 43200, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 64, + "left_input_sampled_relative_l2": 2.1682089914631927e-6, + "n_gauss": 128, + "patch_build_secs": 0.037030641, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 9, + "raw_left_chi": 99, + "raw_left_params": 62240, + "raw_right_chi": 98, + "raw_right_params": 60416, + "rho_max": 8.0, + "right_chi": 64, + "right_input_sampled_relative_l2": 2.8526222942130557e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 0.39476561335676547 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.248353721, + "wall_times_secs": [ + 0.248353721 + ], + "max_error": 3.136094538503786e-6, + "input_max_bond_dim": 64, + "output_max_bond_dim": 64, + "output_bond_dims": [ + 64 + ], + "n_params": 28028, + "n_patches": 1, + "max_patch_bond": 64, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 266189, + "allocated_bytes": 370046427 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-global_fit-chi16.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-global_fit-chi16.json new file mode 100644 index 0000000..915469b --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-global_fit-chi16.json @@ -0,0 +1,64 @@ +{ + "schema_version": 1, + "case": "gaussian_mpo_contraction", + "algorithm": "global_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.10606601717798216, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n2-r6-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004873697, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.006165452, + "input_l2_rtol": 1e-6, + "input_params": 3740, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 16, + "left_input_max_patch_chi": 16, + "left_input_patch_count": 1, + "left_input_patch_params": 2008, + "left_input_sampled_relative_l2": 1.2777851899668411e-6, + "n_gauss": 2, + "patch_build_secs": 0.015757589, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 6, + "raw_left_chi": 32, + "raw_left_params": 4640, + "raw_right_chi": 26, + "raw_right_params": 3872, + "rho_max": 8.0, + "right_chi": 16, + "right_input_max_patch_chi": 16, + "right_input_patch_count": 1, + "right_input_patch_params": 1732, + "right_input_sampled_relative_l2": 9.428700715139774e-7, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.008389418, + "wall_times_secs": [ + 0.008389418 + ], + "max_error": 5.942626396563689e-7, + "input_max_bond_dim": 16, + "output_max_bond_dim": 16, + "output_bond_dims": [ + 4, + 16, + 15, + 10, + 4 + ], + "n_params": 2008, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 38651, + "allocated_bytes": 116757340 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-global_fit-chi24.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-global_fit-chi24.json new file mode 100644 index 0000000..be57592 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-global_fit-chi24.json @@ -0,0 +1,65 @@ +{ + "schema_version": 1, + "case": "gaussian_mpo_contraction", + "algorithm": "global_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.2121320343559643, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n8-r7-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004345993, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.004010053, + "input_l2_rtol": 1e-6, + "input_params": 7924, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 23, + "left_input_max_patch_chi": 23, + "left_input_patch_count": 1, + "left_input_patch_params": 3900, + "left_input_sampled_relative_l2": 1.7758182496310817e-6, + "n_gauss": 8, + "patch_build_secs": 0.035338417, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 7, + "raw_left_chi": 50, + "raw_left_params": 12720, + "raw_right_chi": 50, + "raw_right_params": 12720, + "rho_max": 8.0, + "right_chi": 24, + "right_input_max_patch_chi": 24, + "right_input_patch_count": 1, + "right_input_patch_params": 4024, + "right_input_sampled_relative_l2": 1.4847847441262137e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.015478605, + "wall_times_secs": [ + 0.015478605 + ], + "max_error": 7.31787434896783e-7, + "input_max_bond_dim": 24, + "output_max_bond_dim": 21, + "output_bond_dims": [ + 4, + 16, + 21, + 15, + 10, + 4 + ], + "n_params": 3652, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 45506, + "allocated_bytes": 31268744 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-global_fit-chi41.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-global_fit-chi41.json new file mode 100644 index 0000000..597323a --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-global_fit-chi41.json @@ -0,0 +1,66 @@ +{ + "schema_version": 1, + "case": "gaussian_mpo_contraction", + "algorithm": "global_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.4242640687119286, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n32-r8-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004762783, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.010991831, + "input_l2_rtol": 1e-6, + "input_params": 18380, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 41, + "left_input_max_patch_chi": 41, + "left_input_patch_count": 1, + "left_input_patch_params": 9272, + "left_input_sampled_relative_l2": 1.8640163986326415e-6, + "n_gauss": 32, + "patch_build_secs": 0.050660342, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 8, + "raw_left_chi": 64, + "raw_left_params": 27592, + "raw_right_chi": 64, + "raw_right_params": 27592, + "rho_max": 8.0, + "right_chi": 40, + "right_input_max_patch_chi": 40, + "right_input_patch_count": 1, + "right_input_patch_params": 9108, + "right_input_sampled_relative_l2": 2.232276096287149e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.059542229, + "wall_times_secs": [ + 0.059542229 + ], + "max_error": 1.343038451587278e-6, + "input_max_bond_dim": 41, + "output_max_bond_dim": 34, + "output_bond_dims": [ + 4, + 16, + 34, + 21, + 15, + 10, + 4 + ], + "n_params": 7340, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 60258, + "allocated_bytes": 64712415 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-global_fit-chi64.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-global_fit-chi64.json new file mode 100644 index 0000000..c7d586c --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-global_fit-chi64.json @@ -0,0 +1,67 @@ +{ + "schema_version": 1, + "case": "gaussian_mpo_contraction", + "algorithm": "global_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.8485281374238572, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n128-r9-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.005355165, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.021329536, + "input_l2_rtol": 1e-6, + "input_params": 43200, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 64, + "left_input_max_patch_chi": 64, + "left_input_patch_count": 1, + "left_input_patch_params": 22184, + "left_input_sampled_relative_l2": 2.1682089914631927e-6, + "n_gauss": 128, + "patch_build_secs": 0.070648612, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 9, + "raw_left_chi": 99, + "raw_left_params": 62240, + "raw_right_chi": 98, + "raw_right_params": 60416, + "rho_max": 8.0, + "right_chi": 64, + "right_input_max_patch_chi": 64, + "right_input_patch_count": 1, + "right_input_patch_params": 21016, + "right_input_sampled_relative_l2": 2.8526222942130557e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.119544787, + "wall_times_secs": [ + 0.119544787 + ], + "max_error": 1.5252138234015066e-6, + "input_max_bond_dim": 64, + "output_max_bond_dim": 59, + "output_bond_dims": [ + 4, + 16, + 59, + 36, + 21, + 15, + 10, + 4 + ], + "n_params": 17604, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 77056, + "allocated_bytes": 226336399 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-patched_fit-chi16.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-patched_fit-chi16.json new file mode 100644 index 0000000..e8d8223 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-patched_fit-chi16.json @@ -0,0 +1,63 @@ +{ + "schema_version": 1, + "case": "gaussian_mpo_contraction", + "algorithm": "patched_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.10606601717798216, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n2-r6-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004873697, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.006165452, + "input_l2_rtol": 1e-6, + "input_params": 3740, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 16, + "left_input_max_patch_chi": 16, + "left_input_patch_count": 1, + "left_input_patch_params": 2008, + "left_input_sampled_relative_l2": 1.2777851899668411e-6, + "n_gauss": 2, + "patch_build_secs": 0.015757589, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 6, + "raw_left_chi": 32, + "raw_left_params": 4640, + "raw_right_chi": 26, + "raw_right_params": 3872, + "rho_max": 8.0, + "right_chi": 16, + "right_input_max_patch_chi": 16, + "right_input_patch_count": 1, + "right_input_patch_params": 1732, + "right_input_sampled_relative_l2": 9.428700715139774e-7, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 0.7925205431985778 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.010585742, + "wall_times_secs": [ + 0.010585742 + ], + "max_error": 5.942626405203092e-7, + "input_max_bond_dim": 16, + "output_max_bond_dim": 16, + "output_bond_dims": [ + 16 + ], + "n_params": 2008, + "n_patches": 1, + "max_patch_bond": 16, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 39559, + "allocated_bytes": 11300852 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-patched_fit-chi24.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-patched_fit-chi24.json new file mode 100644 index 0000000..05f353b --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-patched_fit-chi24.json @@ -0,0 +1,63 @@ +{ + "schema_version": 1, + "case": "gaussian_mpo_contraction", + "algorithm": "patched_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.2121320343559643, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n8-r7-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004345993, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.004010053, + "input_l2_rtol": 1e-6, + "input_params": 7924, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 23, + "left_input_max_patch_chi": 23, + "left_input_patch_count": 1, + "left_input_patch_params": 3900, + "left_input_sampled_relative_l2": 1.7758182496310817e-6, + "n_gauss": 8, + "patch_build_secs": 0.035338417, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 7, + "raw_left_chi": 50, + "raw_left_params": 12720, + "raw_right_chi": 50, + "raw_right_params": 12720, + "rho_max": 8.0, + "right_chi": 24, + "right_input_max_patch_chi": 24, + "right_input_patch_count": 1, + "right_input_patch_params": 4024, + "right_input_sampled_relative_l2": 1.4847847441262137e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 0.8413486991511614 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.018397372, + "wall_times_secs": [ + 0.018397372 + ], + "max_error": 7.317874336292172e-7, + "input_max_bond_dim": 24, + "output_max_bond_dim": 21, + "output_bond_dims": [ + 21 + ], + "n_params": 3652, + "n_patches": 1, + "max_patch_bond": 21, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 51035, + "allocated_bytes": 25733817 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-patched_fit-chi41.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-patched_fit-chi41.json new file mode 100644 index 0000000..6d8fbcc --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-patched_fit-chi41.json @@ -0,0 +1,63 @@ +{ + "schema_version": 1, + "case": "gaussian_mpo_contraction", + "algorithm": "patched_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.4242640687119286, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n32-r8-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004762783, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.010991831, + "input_l2_rtol": 1e-6, + "input_params": 18380, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 41, + "left_input_max_patch_chi": 41, + "left_input_patch_count": 1, + "left_input_patch_params": 9272, + "left_input_sampled_relative_l2": 1.8640163986326415e-6, + "n_gauss": 32, + "patch_build_secs": 0.050660342, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 8, + "raw_left_chi": 64, + "raw_left_params": 27592, + "raw_right_chi": 64, + "raw_right_params": 27592, + "rho_max": 8.0, + "right_chi": 40, + "right_input_max_patch_chi": 40, + "right_input_patch_count": 1, + "right_input_patch_params": 9108, + "right_input_sampled_relative_l2": 2.232276096287149e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 1.0846909330428447 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.054893267, + "wall_times_secs": [ + 0.054893267 + ], + "max_error": 1.7733013848135611e-6, + "input_max_bond_dim": 41, + "output_max_bond_dim": 33, + "output_bond_dims": [ + 33 + ], + "n_params": 7192, + "n_patches": 1, + "max_patch_bond": 33, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 71251, + "allocated_bytes": 62648820 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-patched_fit-chi64.json b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-patched_fit-chi64.json new file mode 100644 index 0000000..80f8ad8 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/raw/gaussian_mpo_contraction-patched_fit-chi64.json @@ -0,0 +1,63 @@ +{ + "schema_version": 1, + "case": "gaussian_mpo_contraction", + "algorithm": "patched_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.8485281374238572, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n128-r9-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.005355165, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.021329536, + "input_l2_rtol": 1e-6, + "input_params": 43200, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 64, + "left_input_max_patch_chi": 64, + "left_input_patch_count": 1, + "left_input_patch_params": 22184, + "left_input_sampled_relative_l2": 2.1682089914631927e-6, + "n_gauss": 128, + "patch_build_secs": 0.070648612, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 9, + "raw_left_chi": 99, + "raw_left_params": 62240, + "raw_right_chi": 98, + "raw_right_params": 60416, + "rho_max": 8.0, + "right_chi": 64, + "right_input_max_patch_chi": 64, + "right_input_patch_count": 1, + "right_input_patch_params": 21016, + "right_input_sampled_relative_l2": 2.8526222942130557e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 1.2820271368345622 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.09324669, + "wall_times_secs": [ + 0.09324669 + ], + "max_error": 2.108072680842036e-6, + "input_max_bond_dim": 64, + "output_max_bond_dim": 58, + "output_bond_dims": [ + 58 + ], + "n_params": 16764, + "n_patches": 1, + "max_patch_bond": 58, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 96355, + "allocated_bytes": 158173205 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470-4t/report.md b/result/linux-ryzen-ai-9-hx-470-4t/report.md new file mode 100644 index 0000000..fceebcf --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/report.md @@ -0,0 +1,70 @@ +# Tensor4all benchmark results + +Profile: `linux-ryzen-ai-9-hx-470-4t`. CPU: AMD Ryzen AI 9 HX 470 w/ Radeon 890M. Threads: 4. CPU affinity: 0-3. Source revision: `05ebd78bdc40e09f1534a4bab2ac72131ce271d3`. tensor4all-rs revision: `a5157385f58048355a28a6a003d4cb63d3b1349b`. + +The source revision identifies the clean code that was measured. The commit adding these generated records necessarily follows that revision. + +All timings exclude input construction, cache I/O, format conversion, patch preparation, output conversion and accuracy evaluation. Gaussian inputs use independent two-dimensional interpolative decompositions, balanced pairwise addition, final relative-L2/SVD tolerance `1e-6`, and fixed patch cap 128. + +## Case 1: Fourier elementwise + +| input χ | K | arm | time (s) | sampled relative L2 | output χ | parameters | allocations | allocated (MiB) | +|---:|---:|---|---:|---:|---:|---:|---:|---:| +| 5 | 4 | aci | 0.056474 | 1.166e-08 | 9 | 1026 | 935094 | 92.31 | +| 5 | 4 | fit | 0.025642 | 1.213e-08 | 8 | 606 | 142954 | 23.65 | +| 5 | 4 | naive | 0.001582 | 1.211e-08 | 8 | 606 | 5467 | 2.61 | +| 5 | 4 | zipup | 0.014577 | 1.197e-08 | 8 | 606 | 80520 | 12.11 | +| 8 | 8 | aci | 0.077047 | 8.181e-09 | 16 | 1702 | 1293260 | 129.49 | +| 8 | 8 | fit | 0.018698 | 1.066e-08 | 8 | 726 | 143245 | 24.26 | +| 8 | 8 | naive | 0.001533 | 1.066e-08 | 8 | 726 | 5599 | 3.85 | +| 8 | 8 | zipup | 0.009921 | 1.080e-08 | 8 | 726 | 80806 | 12.58 | +| 8 | 16 | aci | 0.066019 | 6.086e-09 | 16 | 2206 | 1095666 | 111.28 | +| 8 | 16 | fit | 0.019670 | 1.620e-08 | 10 | 900 | 143319 | 24.94 | +| 8 | 16 | naive | 0.002000 | 1.620e-08 | 10 | 900 | 5644 | 5.13 | +| 8 | 16 | zipup | 0.010695 | 1.636e-08 | 10 | 900 | 80869 | 13.09 | +| 11 | 32 | aci | 0.077827 | 4.931e-09 | 23 | 3318 | 1253380 | 136.21 | +| 11 | 32 | fit | 0.019988 | 1.806e-08 | 13 | 1200 | 144030 | 26.34 | +| 11 | 32 | naive | 0.003920 | 1.806e-08 | 13 | 1200 | 6139 | 8.22 | +| 11 | 32 | zipup | 0.011348 | 2.001e-08 | 13 | 1182 | 81544 | 14.05 | +| 14 | 64 | aci | 0.074522 | 4.480e-09 | 29 | 4752 | 1149587 | 129.84 | +| 14 | 64 | fit | 0.021687 | 1.813e-08 | 16 | 1638 | 143911 | 28.08 | +| 14 | 64 | naive | 0.005417 | 1.813e-08 | 16 | 1638 | 6685 | 13.75 | +| 14 | 64 | zipup | 0.011935 | 1.856e-08 | 16 | 1638 | 81873 | 15.28 | +| 16 | 128 | aci | 0.075587 | 3.247e-09 | 32 | 6808 | 1115933 | 135.92 | +| 16 | 128 | fit | 0.023958 | 1.044e-08 | 18 | 2258 | 145567 | 31.98 | +| 16 | 128 | naive | 0.009562 | 1.044e-08 | 18 | 2258 | 7403 | 23.51 | +| 16 | 128 | zipup | 0.016043 | 1.064e-08 | 18 | 2258 | 83392 | 18.25 | + +## Case 2: Gaussian elementwise, global versus patched + +| input χ | raw χ | N | R | arm | time (s) | sampled relative L2 | input patches | input max patch χ | output patches | parameters | allocations | allocated (MiB) | speedup | +|---:|---:|---:|---:|---|---:|---:|---:|---:|---:|---:|---:|---:|---:| +| 16 | 32 | 2 | 6 | global_aci | 0.004899 | 1.398e-06 | 2 | 16 | 1 | 6560 | 57207 | 11.53 | 1.000 | +| 16 | 32 | 2 | 6 | global_fit | 0.010786 | 2.293e-06 | 2 | 16 | 1 | 2300 | 60186 | 114.89 | 1.000 | +| 16 | 32 | 2 | 6 | patched_aci | 0.008488 | 2.292e-06 | 2 | 16 | 1 | 2300 | 81260 | 17.26 | 0.577 | +| 16 | 32 | 2 | 6 | patched_fit | 0.017352 | 2.292e-06 | 2 | 16 | 1 | 2300 | 84856 | 26.06 | 0.622 | +| 24 | 50 | 8 | 7 | global_aci | 0.012319 | 2.942e-06 | 2 | 24 | 1 | 23268 | 87494 | 29.34 | 1.000 | +| 24 | 50 | 8 | 7 | global_fit | 0.017572 | 3.462e-06 | 2 | 24 | 1 | 4776 | 70531 | 26.53 | 1.000 | +| 24 | 50 | 8 | 7 | patched_aci | 0.016877 | 3.459e-06 | 2 | 24 | 1 | 4908 | 116482 | 38.87 | 0.730 | +| 24 | 50 | 8 | 7 | patched_fit | 0.036035 | 3.459e-06 | 2 | 24 | 1 | 4908 | 123259 | 64.21 | 0.488 | +| 41 | 64 | 32 | 8 | global_aci | 0.038988 | 3.468e-06 | 2 | 41 | 1 | 98992 | 110020 | 78.88 | 1.000 | +| 41 | 64 | 32 | 8 | global_fit | 0.042459 | 4.661e-06 | 2 | 41 | 1 | 11072 | 110564 | 73.16 | 1.000 | +| 41 | 64 | 32 | 8 | patched_aci | 0.032272 | 4.366e-06 | 2 | 41 | 1 | 11740 | 117649 | 68.50 | 1.208 | +| 41 | 64 | 32 | 8 | patched_fit | 0.085208 | 4.365e-06 | 2 | 41 | 1 | 11740 | 183635 | 151.13 | 0.498 | +| 64 | 99 | 128 | 9 | global_aci | 0.148520 | 2.705e-06 | 2 | 64 | 1 | 317712 | 141977 | 516.23 | 1.000 | +| 64 | 99 | 128 | 9 | global_fit | 0.098042 | 3.993e-06 | 2 | 64 | 1 | 25476 | 140014 | 175.07 | 1.000 | +| 64 | 99 | 128 | 9 | patched_aci | 0.073028 | 3.187e-06 | 2 | 64 | 1 | 28028 | 150487 | 134.98 | 2.034 | +| 64 | 99 | 128 | 9 | patched_fit | 0.248354 | 3.136e-06 | 2 | 64 | 1 | 28028 | 266189 | 352.90 | 0.395 | + +## Case 3: Gaussian MPO-MPO contraction, global versus patched + +| input χ | raw χ | N | R | arm | time (s) | sampled relative L2 | input patches | input max patch χ | output patches | parameters | allocations | allocated (MiB) | speedup | +|---:|---:|---:|---:|---|---:|---:|---:|---:|---:|---:|---:|---:|---:| +| 16 | 32 | 2 | 6 | global_fit | 0.008389 | 5.943e-07 | 2 | 16 | 1 | 2008 | 38651 | 111.35 | 1.000 | +| 16 | 32 | 2 | 6 | patched_fit | 0.010586 | 5.943e-07 | 2 | 16 | 1 | 2008 | 39559 | 10.78 | 0.793 | +| 24 | 50 | 8 | 7 | global_fit | 0.015479 | 7.318e-07 | 2 | 24 | 1 | 3652 | 45506 | 29.82 | 1.000 | +| 24 | 50 | 8 | 7 | patched_fit | 0.018397 | 7.318e-07 | 2 | 24 | 1 | 3652 | 51035 | 24.54 | 0.841 | +| 41 | 64 | 32 | 8 | global_fit | 0.059542 | 1.343e-06 | 2 | 41 | 1 | 7340 | 60258 | 61.71 | 1.000 | +| 41 | 64 | 32 | 8 | patched_fit | 0.054893 | 1.773e-06 | 2 | 41 | 1 | 7192 | 71251 | 59.75 | 1.085 | +| 64 | 99 | 128 | 9 | global_fit | 0.119545 | 1.525e-06 | 2 | 64 | 1 | 17604 | 77056 | 215.85 | 1.000 | +| 64 | 99 | 128 | 9 | patched_fit | 0.093247 | 2.108e-06 | 2 | 64 | 1 | 16764 | 96355 | 150.85 | 1.282 | diff --git a/result/linux-ryzen-ai-9-hx-470-4t/run.yaml b/result/linux-ryzen-ai-9-hx-470-4t/run.yaml new file mode 100644 index 0000000..e3b1ba0 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470-4t/run.yaml @@ -0,0 +1,9 @@ +profile: linux-ryzen-ai-9-hx-470-4t +date_utc: 2026-10-10T23:07:58Z +os: Linux +chip: AMD Ryzen AI 9 HX 470 w/ Radeon 890M +memory_gb: 58.5 +repo_rev: 05ebd78bdc40e09f1534a4bab2ac72131ce271d3 +tensor4all_rs_rev: a5157385f58048355a28a6a003d4cb63d3b1349b +threads: 4 +cpu_affinity: 0-3 diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k128.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k128.json new file mode 100644 index 0000000..b94d9bc --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k128.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "aci", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "aci_absolute_residual", + "k_max": 128, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.071914685, + "wall_times_secs": [ + 0.075926434, + 0.072072008, + 0.071914685, + 0.07182632, + 0.071762901 + ], + "max_error": 3.2468908083116903e-9, + "input_max_bond_dim": 16, + "output_max_bond_dim": 32, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 32, + 30, + 24, + 17, + 14, + 10, + 8, + 6, + 6, + 4, + 4, + 4, + 4, + 3, + 2 + ], + "n_params": 6808, + "n_allocations": 1115945, + "allocated_bytes": 142567242 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k16.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k16.json new file mode 100644 index 0000000..c830034 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k16.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "aci", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "aci_absolute_residual", + "k_max": 16, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.063716851, + "wall_times_secs": [ + 0.063716851, + 0.063504194, + 0.063649966, + 0.063779679, + 0.064860944 + ], + "max_error": 6.085556560690714e-9, + "input_max_bond_dim": 8, + "output_max_bond_dim": 16, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 16, + 13, + 9, + 9, + 7, + 7, + 5, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 2 + ], + "n_params": 2206, + "n_allocations": 1095665, + "allocated_bytes": 116685946 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k32.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k32.json new file mode 100644 index 0000000..54bd792 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k32.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "aci", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "aci_absolute_residual", + "k_max": 32, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.074268954, + "wall_times_secs": [ + 0.074268954, + 0.0740416, + 0.074473155, + 0.074301235, + 0.074150242 + ], + "max_error": 4.930920676253506e-9, + "input_max_bond_dim": 11, + "output_max_bond_dim": 23, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 23, + 17, + 14, + 10, + 9, + 7, + 7, + 5, + 5, + 5, + 4, + 4, + 3, + 3, + 2 + ], + "n_params": 3318, + "n_allocations": 1253379, + "allocated_bytes": 142823569 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k4.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k4.json new file mode 100644 index 0000000..d51b1ae --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k4.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "aci", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "aci_absolute_residual", + "k_max": 4, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.051510624, + "wall_times_secs": [ + 0.051510624, + 0.051383929, + 0.051589852, + 0.0515507, + 0.051285865 + ], + "max_error": 1.1659684028740639e-8, + "input_max_bond_dim": 5, + "output_max_bond_dim": 9, + "output_bond_dims": [ + 2, + 4, + 8, + 9, + 8, + 8, + 7, + 7, + 5, + 5, + 4, + 4, + 4, + 3, + 3, + 3, + 2, + 2, + 2 + ], + "n_params": 1026, + "n_allocations": 935093, + "allocated_bytes": 96796530 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k64.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k64.json new file mode 100644 index 0000000..ae1b225 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k64.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "aci", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "aci_absolute_residual", + "k_max": 64, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.070777474, + "wall_times_secs": [ + 0.071094295, + 0.070431179, + 0.070777474, + 0.070740195, + 0.07081282 + ], + "max_error": 4.479700684341785e-9, + "input_max_bond_dim": 14, + "output_max_bond_dim": 29, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 29, + 22, + 17, + 14, + 10, + 9, + 7, + 7, + 5, + 5, + 5, + 5, + 4, + 3, + 2 + ], + "n_params": 4752, + "n_allocations": 1149586, + "allocated_bytes": 136140887 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k8.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k8.json new file mode 100644 index 0000000..b13ffae --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-aci-k8.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "aci", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "aci_absolute_residual", + "k_max": 8, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.07372522, + "wall_times_secs": [ + 0.073480665, + 0.073351393, + 0.07372522, + 0.08029019, + 0.076624686 + ], + "max_error": 8.181184498521454e-9, + "input_max_bond_dim": 8, + "output_max_bond_dim": 16, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 13, + 9, + 9, + 7, + 7, + 5, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 2, + 2 + ], + "n_params": 1702, + "n_allocations": 1293258, + "allocated_bytes": 135778144 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k128.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k128.json new file mode 100644 index 0000000..93fcc66 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k128.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "fit", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 128, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.022184981, + "wall_times_secs": [ + 0.022780741, + 0.021999425, + 0.022230896, + 0.022184981, + 0.021219891 + ], + "max_error": 1.044043662371722e-8, + "input_max_bond_dim": 16, + "output_max_bond_dim": 18, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 18, + 13, + 10, + 8, + 7, + 6, + 5, + 4, + 4, + 4, + 3, + 3, + 3, + 3, + 2 + ], + "n_params": 2258, + "n_allocations": 143793, + "allocated_bytes": 33462902 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k16.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k16.json new file mode 100644 index 0000000..6edc6e9 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k16.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "fit", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 16, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.013237359, + "wall_times_secs": [ + 0.013320444, + 0.013237359, + 0.013154956, + 0.013170906, + 0.014282938 + ], + "max_error": 1.6203459440631483e-8, + "input_max_bond_dim": 8, + "output_max_bond_dim": 10, + "output_bond_dims": [ + 2, + 4, + 8, + 10, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2 + ], + "n_params": 900, + "n_allocations": 142989, + "allocated_bytes": 26115794 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k32.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k32.json new file mode 100644 index 0000000..14e8e1d --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k32.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "fit", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 32, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.013954286, + "wall_times_secs": [ + 0.014256179, + 0.014014047, + 0.013900305, + 0.013892631, + 0.013954286 + ], + "max_error": 1.8064613036018543e-8, + "input_max_bond_dim": 11, + "output_max_bond_dim": 13, + "output_bond_dims": [ + 2, + 4, + 8, + 13, + 10, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2, + 2 + ], + "n_params": 1200, + "n_allocations": 143389, + "allocated_bytes": 27571434 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k4.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k4.json new file mode 100644 index 0000000..a759510 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k4.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "fit", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 4, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.012444862, + "wall_times_secs": [ + 0.012513269, + 0.012504042, + 0.012444862, + 0.012348312, + 0.012305462 + ], + "max_error": 1.2131007131617917e-8, + "input_max_bond_dim": 5, + "output_max_bond_dim": 8, + "output_bond_dims": [ + 2, + 4, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2, + 2 + ], + "n_params": 606, + "n_allocations": 142866, + "allocated_bytes": 24746157 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k64.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k64.json new file mode 100644 index 0000000..a9c0516 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k64.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "fit", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 64, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.014982423, + "wall_times_secs": [ + 0.017306084, + 0.01487385, + 0.014914476, + 0.015369174, + 0.014982423 + ], + "max_error": 1.8126118528847354e-8, + "input_max_bond_dim": 14, + "output_max_bond_dim": 16, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 13, + 10, + 8, + 7, + 6, + 5, + 4, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2 + ], + "n_params": 1638, + "n_allocations": 143592, + "allocated_bytes": 29413245 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k8.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k8.json new file mode 100644 index 0000000..d0cc5fe --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-fit-k8.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "fit", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 8, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.01295367, + "wall_times_secs": [ + 0.01295367, + 0.013217783, + 0.013015255, + 0.012831793, + 0.012826222 + ], + "max_error": 1.0663293834392334e-8, + "input_max_bond_dim": 8, + "output_max_bond_dim": 8, + "output_bond_dims": [ + 2, + 4, + 8, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2 + ], + "n_params": 726, + "n_allocations": 143003, + "allocated_bytes": 25400069 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k128.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k128.json new file mode 100644 index 0000000..3d48d67 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k128.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "naive", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 128, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.011175225, + "wall_times_secs": [ + 0.009191626, + 0.009144598, + 0.011702157, + 0.011175225, + 0.011252048 + ], + "max_error": 1.0441368106897123e-8, + "input_max_bond_dim": 16, + "output_max_bond_dim": 18, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 18, + 13, + 10, + 8, + 7, + 6, + 5, + 4, + 4, + 4, + 3, + 3, + 3, + 3, + 2 + ], + "n_params": 2258, + "n_allocations": 5374, + "allocated_bytes": 24589164 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k16.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k16.json new file mode 100644 index 0000000..394e2ee --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k16.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "naive", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 16, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.001710208, + "wall_times_secs": [ + 0.001768938, + 0.001722021, + 0.001710208, + 0.001697514, + 0.001704738 + ], + "max_error": 1.62021859796595e-8, + "input_max_bond_dim": 8, + "output_max_bond_dim": 10, + "output_bond_dims": [ + 2, + 4, + 8, + 10, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2 + ], + "n_params": 900, + "n_allocations": 5323, + "allocated_bytes": 5369132 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k32.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k32.json new file mode 100644 index 0000000..07ef732 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k32.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "naive", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 32, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.002830036, + "wall_times_secs": [ + 0.003734593, + 0.002857277, + 0.002830036, + 0.002824266, + 0.002813425 + ], + "max_error": 1.806346085824233e-8, + "input_max_bond_dim": 11, + "output_max_bond_dim": 13, + "output_bond_dims": [ + 2, + 4, + 8, + 13, + 10, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2, + 2 + ], + "n_params": 1200, + "n_allocations": 5346, + "allocated_bytes": 8600788 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k4.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k4.json new file mode 100644 index 0000000..d8a1a9a --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k4.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "naive", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 4, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.000917511, + "wall_times_secs": [ + 0.000929483, + 0.000917511, + 0.000915476, + 0.000925545, + 0.00090119 + ], + "max_error": 1.2110321672966727e-8, + "input_max_bond_dim": 5, + "output_max_bond_dim": 8, + "output_bond_dims": [ + 2, + 4, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2, + 2 + ], + "n_params": 606, + "n_allocations": 5323, + "allocated_bytes": 2733108 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k64.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k64.json new file mode 100644 index 0000000..5353d8c --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k64.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "naive", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 64, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.004956561, + "wall_times_secs": [ + 0.004997177, + 0.004971458, + 0.004954657, + 0.004940962, + 0.004956561 + ], + "max_error": 1.8125685779572552e-8, + "input_max_bond_dim": 14, + "output_max_bond_dim": 16, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 13, + 10, + 8, + 7, + 6, + 5, + 4, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2 + ], + "n_params": 1638, + "n_allocations": 5363, + "allocated_bytes": 14381948 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k8.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k8.json new file mode 100644 index 0000000..3e5ccb3 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-naive-k8.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "naive", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 8, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.00131956, + "wall_times_secs": [ + 0.001327115, + 0.00131956, + 0.001302178, + 0.001417713, + 0.001305735 + ], + "max_error": 1.0657100928094615e-8, + "input_max_bond_dim": 8, + "output_max_bond_dim": 8, + "output_bond_dims": [ + 2, + 4, + 8, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2 + ], + "n_params": 726, + "n_allocations": 5323, + "allocated_bytes": 4033316 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k128.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k128.json new file mode 100644 index 0000000..063e474 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k128.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "zipup", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 128, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.010941228, + "wall_times_secs": [ + 0.01087246, + 0.010667979, + 0.010941228, + 0.011391999, + 0.011257628 + ], + "max_error": 1.063887427760653e-8, + "input_max_bond_dim": 16, + "output_max_bond_dim": 18, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 18, + 13, + 10, + 8, + 7, + 6, + 5, + 4, + 4, + 4, + 3, + 3, + 3, + 3, + 2 + ], + "n_params": 2258, + "n_allocations": 81754, + "allocated_bytes": 19081338 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k16.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k16.json new file mode 100644 index 0000000..691d388 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k16.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "zipup", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 16, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.006922566, + "wall_times_secs": [ + 0.007071263, + 0.006994641, + 0.006915593, + 0.006922566, + 0.006920933 + ], + "max_error": 1.636334768856949e-8, + "input_max_bond_dim": 8, + "output_max_bond_dim": 10, + "output_bond_dims": [ + 2, + 4, + 8, + 10, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2 + ], + "n_params": 900, + "n_allocations": 80547, + "allocated_bytes": 13665296 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k32.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k32.json new file mode 100644 index 0000000..16828e8 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k32.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "zipup", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 32, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.007408692, + "wall_times_secs": [ + 0.007408692, + 0.007342779, + 0.00736405, + 0.008155996, + 0.007431936 + ], + "max_error": 2.0007564441816385e-8, + "input_max_bond_dim": 11, + "output_max_bond_dim": 13, + "output_bond_dims": [ + 2, + 4, + 8, + 13, + 10, + 8, + 7, + 6, + 5, + 4, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2, + 2 + ], + "n_params": 1182, + "n_allocations": 80911, + "allocated_bytes": 14700272 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k4.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k4.json new file mode 100644 index 0000000..4d68580 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k4.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "zipup", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 4, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.006449224, + "wall_times_secs": [ + 0.00680674, + 0.006529443, + 0.006449224, + 0.0064255, + 0.006446008 + ], + "max_error": 1.1974477610213059e-8, + "input_max_bond_dim": 5, + "output_max_bond_dim": 8, + "output_bond_dims": [ + 2, + 4, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2, + 2 + ], + "n_params": 606, + "n_allocations": 80441, + "allocated_bytes": 12674939 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k64.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k64.json new file mode 100644 index 0000000..3c654a8 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k64.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "zipup", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 64, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.007747985, + "wall_times_secs": [ + 0.007764015, + 0.007747985, + 0.007710595, + 0.007717569, + 0.007763714 + ], + "max_error": 1.8557199283787496e-8, + "input_max_bond_dim": 14, + "output_max_bond_dim": 16, + "output_bond_dims": [ + 2, + 4, + 8, + 16, + 13, + 10, + 8, + 7, + 6, + 5, + 4, + 4, + 4, + 3, + 3, + 3, + 3, + 2, + 2 + ], + "n_params": 1638, + "n_allocations": 81012, + "allocated_bytes": 15986938 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k8.json b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k8.json new file mode 100644 index 0000000..1fa126c --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/elementwise_fourier-zipup-k8.json @@ -0,0 +1,54 @@ +{ + "schema_version": 1, + "case": "elementwise_fourier", + "algorithm": "zipup", + "params": { + "error_metric": "sampled_relative_l2", + "error_seed": 999, + "fit_nfullsweeps": 2, + "internal_tolerance_metric": "relative_l2_svd", + "k_max": 8, + "max_bond": 4096, + "n_error_samples": 256, + "r": 20, + "runs": 5, + "warmups": 1 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.006675756, + "wall_times_secs": [ + 0.006740698, + 0.007586864, + 0.006675756, + 0.006649627, + 0.006646231 + ], + "max_error": 1.0797260594548758e-8, + "input_max_bond_dim": 8, + "output_max_bond_dim": 8, + "output_bond_dims": [ + 2, + 4, + 8, + 8, + 7, + 6, + 5, + 5, + 4, + 4, + 3, + 3, + 3, + 3, + 3, + 2, + 2, + 2, + 2 + ], + "n_params": 726, + "n_allocations": 80573, + "allocated_bytes": 13167091 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_aci-chi16.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_aci-chi16.json new file mode 100644 index 0000000..017bea3 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_aci-chi16.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "global_aci", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.10606601717798216, + "cache_hit": false, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n2-r6-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.0, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 2.293705879, + "input_compression_secs": 0.001028978, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 16, + "input_params": 3740, + "input_patch_count": 2, + "input_patch_params": 3740, + "internal_tolerance": 1e-8, + "internal_tolerance_metric": "aci_scale_relative_residual", + "interpolation_tolerance": 1e-10, + "left_chi": 16, + "left_input_sampled_relative_l2": 1.277785189967113e-6, + "n_gauss": 2, + "patch_build_secs": 0.009627128, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 6, + "raw_left_chi": 32, + "raw_left_params": 4640, + "raw_right_chi": 26, + "raw_right_params": 3872, + "rho_max": 8.0, + "right_chi": 16, + "right_input_sampled_relative_l2": 9.428700715139774e-7, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.004388421, + "wall_times_secs": [ + 0.004388421 + ], + "max_error": 1.398200455696646e-6, + "input_max_bond_dim": 16, + "output_max_bond_dim": 47, + "output_bond_dims": [ + 4, + 16, + 47, + 16, + 4 + ], + "n_params": 6560, + "input_build_secs": 2.293705879, + "n_allocations": 57157, + "allocated_bytes": 12012948 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_aci-chi24.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_aci-chi24.json new file mode 100644 index 0000000..57b9679 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_aci-chi24.json @@ -0,0 +1,61 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "global_aci", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.2121320343559643, + "cache_hit": false, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n8-r7-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.0, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 17.225256878, + "input_compression_secs": 0.002320416, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 24, + "input_params": 7924, + "input_patch_count": 2, + "input_patch_params": 7924, + "internal_tolerance": 1e-8, + "internal_tolerance_metric": "aci_scale_relative_residual", + "interpolation_tolerance": 1e-10, + "left_chi": 23, + "left_input_sampled_relative_l2": 1.7758182495895119e-6, + "n_gauss": 8, + "patch_build_secs": 0.012837342, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 7, + "raw_left_chi": 50, + "raw_left_params": 12720, + "raw_right_chi": 50, + "raw_right_params": 12720, + "rho_max": 8.0, + "right_chi": 24, + "right_input_sampled_relative_l2": 1.4847847441728126e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.0114308, + "wall_times_secs": [ + 0.0114308 + ], + "max_error": 2.9417902475304512e-6, + "input_max_bond_dim": 24, + "output_max_bond_dim": 64, + "output_bond_dims": [ + 4, + 16, + 64, + 59, + 15, + 4 + ], + "n_params": 23268, + "input_build_secs": 17.225256878, + "n_allocations": 90766, + "allocated_bytes": 31137116 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_aci-chi41.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_aci-chi41.json new file mode 100644 index 0000000..39d5f28 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_aci-chi41.json @@ -0,0 +1,62 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "global_aci", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.4242640687119286, + "cache_hit": false, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n32-r8-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.0, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 105.754228663, + "input_compression_secs": 0.004790109, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 41, + "input_params": 18380, + "input_patch_count": 2, + "input_patch_params": 18380, + "internal_tolerance": 1e-8, + "internal_tolerance_metric": "aci_scale_relative_residual", + "interpolation_tolerance": 1e-10, + "left_chi": 41, + "left_input_sampled_relative_l2": 1.864016398611087e-6, + "n_gauss": 32, + "patch_build_secs": 0.017721563, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 8, + "raw_left_chi": 64, + "raw_left_params": 27592, + "raw_right_chi": 64, + "raw_right_params": 27592, + "rho_max": 8.0, + "right_chi": 40, + "right_input_sampled_relative_l2": 2.2322760961784926e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.034805899, + "wall_times_secs": [ + 0.034805899 + ], + "max_error": 3.4680690587613565e-6, + "input_max_bond_dim": 41, + "output_max_bond_dim": 183, + "output_bond_dims": [ + 4, + 16, + 64, + 183, + 60, + 15, + 4 + ], + "n_params": 98992, + "input_build_secs": 105.754228663, + "n_allocations": 107105, + "allocated_bytes": 79198084 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_aci-chi64.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_aci-chi64.json new file mode 100644 index 0000000..32b2e28 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_aci-chi64.json @@ -0,0 +1,63 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "global_aci", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.8485281374238572, + "cache_hit": false, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n128-r9-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.0, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 583.614199605, + "input_compression_secs": 0.010811515, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 64, + "input_params": 43200, + "input_patch_count": 2, + "input_patch_params": 43200, + "internal_tolerance": 1e-8, + "internal_tolerance_metric": "aci_scale_relative_residual", + "interpolation_tolerance": 1e-10, + "left_chi": 64, + "left_input_sampled_relative_l2": 2.168208991570596e-6, + "n_gauss": 128, + "patch_build_secs": 0.028485206, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 9, + "raw_left_chi": 99, + "raw_left_params": 62240, + "raw_right_chi": 98, + "raw_right_params": 60416, + "rho_max": 8.0, + "right_chi": 64, + "right_input_sampled_relative_l2": 2.8526222945123154e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.149108338, + "wall_times_secs": [ + 0.149108338 + ], + "max_error": 2.704573944278199e-6, + "input_max_bond_dim": 64, + "output_max_bond_dim": 256, + "output_bond_dims": [ + 4, + 16, + 64, + 256, + 193, + 60, + 15, + 4 + ], + "n_params": 317712, + "input_build_secs": 583.614199605, + "n_allocations": 139715, + "allocated_bytes": 237010164 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_fit-chi16.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_fit-chi16.json new file mode 100644 index 0000000..fc3836a --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_fit-chi16.json @@ -0,0 +1,61 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "global_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.10606601717798216, + "cache_hit": false, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n2-r6-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.0, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 2.293705879, + "input_compression_secs": 0.001028978, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 16, + "input_params": 3740, + "input_patch_count": 2, + "input_patch_params": 3740, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 16, + "left_input_sampled_relative_l2": 1.277785189967113e-6, + "n_gauss": 2, + "patch_build_secs": 0.009627128, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 6, + "raw_left_chi": 32, + "raw_left_params": 4640, + "raw_right_chi": 26, + "raw_right_params": 3872, + "rho_max": 8.0, + "right_chi": 16, + "right_input_sampled_relative_l2": 9.428700715139774e-7, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.007956603, + "wall_times_secs": [ + 0.007956603 + ], + "max_error": 2.293034800180824e-6, + "input_max_bond_dim": 16, + "output_max_bond_dim": 17, + "output_bond_dims": [ + 4, + 16, + 17, + 11, + 4 + ], + "n_params": 2300, + "rtol": 1e-6, + "input_build_secs": 2.293705879, + "n_allocations": 57856, + "allocated_bytes": 13660644 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_fit-chi24.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_fit-chi24.json new file mode 100644 index 0000000..ff5b175 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_fit-chi24.json @@ -0,0 +1,62 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "global_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.2121320343559643, + "cache_hit": false, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n8-r7-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.0, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 17.225256878, + "input_compression_secs": 0.002320416, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 24, + "input_params": 7924, + "input_patch_count": 2, + "input_patch_params": 7924, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 23, + "left_input_sampled_relative_l2": 1.7758182495895119e-6, + "n_gauss": 8, + "patch_build_secs": 0.012837342, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 7, + "raw_left_chi": 50, + "raw_left_params": 12720, + "raw_right_chi": 50, + "raw_right_params": 12720, + "rho_max": 8.0, + "right_chi": 24, + "right_input_sampled_relative_l2": 1.4847847441728126e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.013576099, + "wall_times_secs": [ + 0.013576099 + ], + "max_error": 3.4615301396861987e-6, + "input_max_bond_dim": 24, + "output_max_bond_dim": 27, + "output_bond_dims": [ + 4, + 16, + 27, + 17, + 11, + 4 + ], + "n_params": 4776, + "rtol": 1e-6, + "input_build_secs": 17.225256878, + "n_allocations": 65431, + "allocated_bytes": 23457611 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_fit-chi41.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_fit-chi41.json new file mode 100644 index 0000000..28be913 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_fit-chi41.json @@ -0,0 +1,63 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "global_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.4242640687119286, + "cache_hit": false, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n32-r8-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.0, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 105.754228663, + "input_compression_secs": 0.004790109, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 41, + "input_params": 18380, + "input_patch_count": 2, + "input_patch_params": 18380, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 41, + "left_input_sampled_relative_l2": 1.864016398611087e-6, + "n_gauss": 32, + "patch_build_secs": 0.017721563, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 8, + "raw_left_chi": 64, + "raw_left_params": 27592, + "raw_right_chi": 64, + "raw_right_params": 27592, + "rho_max": 8.0, + "right_chi": 40, + "right_input_sampled_relative_l2": 2.2322760961784926e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.040617822, + "wall_times_secs": [ + 0.040617822 + ], + "max_error": 4.661153609486412e-6, + "input_max_bond_dim": 41, + "output_max_bond_dim": 48, + "output_bond_dims": [ + 4, + 16, + 48, + 27, + 16, + 10, + 4 + ], + "n_params": 11072, + "rtol": 1e-6, + "input_build_secs": 105.754228663, + "n_allocations": 99690, + "allocated_bytes": 59602118 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_fit-chi64.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_fit-chi64.json new file mode 100644 index 0000000..4ea5ce0 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-global_fit-chi64.json @@ -0,0 +1,64 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "global_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.8485281374238572, + "cache_hit": false, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n128-r9-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.0, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 583.614199605, + "input_compression_secs": 0.010811515, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 64, + "input_params": 43200, + "input_patch_count": 2, + "input_patch_params": 43200, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 64, + "left_input_sampled_relative_l2": 2.168208991570596e-6, + "n_gauss": 128, + "patch_build_secs": 0.028485206, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 9, + "raw_left_chi": 99, + "raw_left_params": 62240, + "raw_right_chi": 98, + "raw_right_params": 60416, + "rho_max": 8.0, + "right_chi": 64, + "right_input_sampled_relative_l2": 2.8526222945123154e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.15366022, + "wall_times_secs": [ + 0.15366022 + ], + "max_error": 3.9932681897969615e-6, + "input_max_bond_dim": 64, + "output_max_bond_dim": 64, + "output_bond_dims": [ + 4, + 16, + 64, + 51, + 27, + 16, + 10, + 4 + ], + "n_params": 25476, + "rtol": 1e-6, + "input_build_secs": 583.614199605, + "n_allocations": 110346, + "allocated_bytes": 182769532 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_aci-chi16.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_aci-chi16.json new file mode 100644 index 0000000..f6f92d4 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_aci-chi16.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "patched_aci", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.10606601717798216, + "cache_hit": false, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n2-r6-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.0, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 2.293705879, + "input_compression_secs": 0.001028978, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 16, + "input_params": 3740, + "input_patch_count": 2, + "input_patch_params": 3740, + "internal_tolerance": 1e-8, + "internal_tolerance_metric": "aci_absolute_patch_residual_then_global_l2", + "interpolation_tolerance": 1e-10, + "left_chi": 16, + "left_input_sampled_relative_l2": 1.277785189967113e-6, + "n_gauss": 2, + "patch_build_secs": 0.009627128, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 6, + "raw_left_chi": 32, + "raw_left_params": 4640, + "raw_right_chi": 26, + "raw_right_params": 3872, + "rho_max": 8.0, + "right_chi": 16, + "right_input_sampled_relative_l2": 9.428700715139774e-7, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 0.6223044980726435 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.007051887, + "wall_times_secs": [ + 0.007051887 + ], + "max_error": 2.291749915052518e-6, + "input_max_bond_dim": 16, + "output_max_bond_dim": 17, + "output_bond_dims": [ + 17 + ], + "n_params": 2300, + "n_patches": 1, + "max_patch_bond": 17, + "rtol": 1e-6, + "input_build_secs": 2.293705879, + "n_allocations": 82738, + "allocated_bytes": 18491600 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_aci-chi24.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_aci-chi24.json new file mode 100644 index 0000000..3329090 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_aci-chi24.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "patched_aci", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.2121320343559643, + "cache_hit": false, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n8-r7-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.0, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 17.225256878, + "input_compression_secs": 0.002320416, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 24, + "input_params": 7924, + "input_patch_count": 2, + "input_patch_params": 7924, + "internal_tolerance": 1e-8, + "internal_tolerance_metric": "aci_absolute_patch_residual_then_global_l2", + "interpolation_tolerance": 1e-10, + "left_chi": 23, + "left_input_sampled_relative_l2": 1.7758182495895119e-6, + "n_gauss": 8, + "patch_build_secs": 0.012837342, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 7, + "raw_left_chi": 50, + "raw_left_params": 12720, + "raw_right_chi": 50, + "raw_right_params": 12720, + "rho_max": 8.0, + "right_chi": 24, + "right_input_sampled_relative_l2": 1.4847847441728126e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 0.7676577466557515 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.01489049, + "wall_times_secs": [ + 0.01489049 + ], + "max_error": 3.4589838847219463e-6, + "input_max_bond_dim": 24, + "output_max_bond_dim": 28, + "output_bond_dims": [ + 28 + ], + "n_params": 4908, + "n_patches": 1, + "max_patch_bond": 28, + "rtol": 1e-6, + "input_build_secs": 17.225256878, + "n_allocations": 117462, + "allocated_bytes": 41356287 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_aci-chi41.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_aci-chi41.json new file mode 100644 index 0000000..ac7adb8 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_aci-chi41.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "patched_aci", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.4242640687119286, + "cache_hit": false, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n32-r8-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.0, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 105.754228663, + "input_compression_secs": 0.004790109, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 41, + "input_params": 18380, + "input_patch_count": 2, + "input_patch_params": 18380, + "internal_tolerance": 1e-8, + "internal_tolerance_metric": "aci_absolute_patch_residual_then_global_l2", + "interpolation_tolerance": 1e-10, + "left_chi": 41, + "left_input_sampled_relative_l2": 1.864016398611087e-6, + "n_gauss": 32, + "patch_build_secs": 0.017721563, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 8, + "raw_left_chi": 64, + "raw_left_params": 27592, + "raw_right_chi": 64, + "raw_right_params": 27592, + "rho_max": 8.0, + "right_chi": 40, + "right_input_sampled_relative_l2": 2.2322760961784926e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 1.22131566599618 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.028498692, + "wall_times_secs": [ + 0.028498692 + ], + "max_error": 4.3658796754637735e-6, + "input_max_bond_dim": 41, + "output_max_bond_dim": 48, + "output_bond_dims": [ + 48 + ], + "n_params": 11740, + "n_patches": 1, + "max_patch_bond": 48, + "rtol": 1e-6, + "input_build_secs": 105.754228663, + "n_allocations": 111749, + "allocated_bytes": 71649121 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_aci-chi64.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_aci-chi64.json new file mode 100644 index 0000000..cf2d12f --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_aci-chi64.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "patched_aci", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.8485281374238572, + "cache_hit": false, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n128-r9-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.0, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 583.614199605, + "input_compression_secs": 0.010811515, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 64, + "input_params": 43200, + "input_patch_count": 2, + "input_patch_params": 43200, + "internal_tolerance": 1e-8, + "internal_tolerance_metric": "aci_absolute_patch_residual_then_global_l2", + "interpolation_tolerance": 1e-10, + "left_chi": 64, + "left_input_sampled_relative_l2": 2.168208991570596e-6, + "n_gauss": 128, + "patch_build_secs": 0.028485206, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 9, + "raw_left_chi": 99, + "raw_left_params": 62240, + "raw_right_chi": 98, + "raw_right_params": 60416, + "rho_max": 8.0, + "right_chi": 64, + "right_input_sampled_relative_l2": 2.8526222945123154e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 2.1455310461486268 + }, + "seed": 0, + "tolerance": 1e-8, + "wall_time_median_secs": 0.069497171, + "wall_times_secs": [ + 0.069497171 + ], + "max_error": 3.1786362740241002e-6, + "input_max_bond_dim": 64, + "output_max_bond_dim": 64, + "output_bond_dims": [ + 64 + ], + "n_params": 28028, + "n_patches": 1, + "max_patch_bond": 64, + "rtol": 1e-6, + "input_build_secs": 583.614199605, + "n_allocations": 140133, + "allocated_bytes": 141210371 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_fit-chi16.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_fit-chi16.json new file mode 100644 index 0000000..39932be --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_fit-chi16.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "patched_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.10606601717798216, + "cache_hit": false, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n2-r6-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.0, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 2.293705879, + "input_compression_secs": 0.001028978, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 16, + "input_params": 3740, + "input_patch_count": 2, + "input_patch_params": 3740, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 16, + "left_input_sampled_relative_l2": 1.277785189967113e-6, + "n_gauss": 2, + "patch_build_secs": 0.009627128, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 6, + "raw_left_chi": 32, + "raw_left_params": 4640, + "raw_right_chi": 26, + "raw_right_params": 3872, + "rho_max": 8.0, + "right_chi": 16, + "right_input_sampled_relative_l2": 9.428700715139774e-7, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 0.5816271079095292 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.013679904, + "wall_times_secs": [ + 0.013679904 + ], + "max_error": 2.2917390950832613e-6, + "input_max_bond_dim": 16, + "output_max_bond_dim": 17, + "output_bond_dims": [ + 17 + ], + "n_params": 2300, + "n_patches": 1, + "max_patch_bond": 17, + "rtol": 1e-6, + "input_build_secs": 2.293705879, + "n_allocations": 82983, + "allocated_bytes": 25197056 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_fit-chi24.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_fit-chi24.json new file mode 100644 index 0000000..d32304a --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_fit-chi24.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "patched_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.2121320343559643, + "cache_hit": false, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n8-r7-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.0, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 17.225256878, + "input_compression_secs": 0.002320416, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 24, + "input_params": 7924, + "input_patch_count": 2, + "input_patch_params": 7924, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 23, + "left_input_sampled_relative_l2": 1.7758182495895119e-6, + "n_gauss": 8, + "patch_build_secs": 0.012837342, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 7, + "raw_left_chi": 50, + "raw_left_params": 12720, + "raw_right_chi": 50, + "raw_right_params": 12720, + "rho_max": 8.0, + "right_chi": 24, + "right_input_sampled_relative_l2": 1.4847847441728126e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 0.40641850970149473 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.033404234, + "wall_times_secs": [ + 0.033404234 + ], + "max_error": 3.4589720252190003e-6, + "input_max_bond_dim": 24, + "output_max_bond_dim": 28, + "output_bond_dims": [ + 28 + ], + "n_params": 4908, + "n_patches": 1, + "max_patch_bond": 28, + "rtol": 1e-6, + "input_build_secs": 17.225256878, + "n_allocations": 110716, + "allocated_bytes": 58033781 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_fit-chi41.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_fit-chi41.json new file mode 100644 index 0000000..2d0c527 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_fit-chi41.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "patched_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.4242640687119286, + "cache_hit": false, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n32-r8-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.0, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 105.754228663, + "input_compression_secs": 0.004790109, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 41, + "input_params": 18380, + "input_patch_count": 2, + "input_patch_params": 18380, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 41, + "left_input_sampled_relative_l2": 1.864016398611087e-6, + "n_gauss": 32, + "patch_build_secs": 0.017721563, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 8, + "raw_left_chi": 64, + "raw_left_params": 27592, + "raw_right_chi": 64, + "raw_right_params": 27592, + "rho_max": 8.0, + "right_chi": 40, + "right_input_sampled_relative_l2": 2.2322760961784926e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 0.3817843351578471 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.106389441, + "wall_times_secs": [ + 0.106389441 + ], + "max_error": 4.3654385061452385e-6, + "input_max_bond_dim": 41, + "output_max_bond_dim": 48, + "output_bond_dims": [ + 48 + ], + "n_params": 11740, + "n_patches": 1, + "max_patch_bond": 48, + "rtol": 1e-6, + "input_build_secs": 105.754228663, + "n_allocations": 157617, + "allocated_bytes": 156738731 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_fit-chi64.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_fit-chi64.json new file mode 100644 index 0000000..1db4e28 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_elementwise-patched_fit-chi64.json @@ -0,0 +1,60 @@ +{ + "schema_version": 1, + "case": "gaussian_elementwise", + "algorithm": "patched_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.8485281374238572, + "cache_hit": false, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n128-r9-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.0, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 583.614199605, + "input_compression_secs": 0.010811515, + "input_l2_rtol": 1e-6, + "input_max_patch_chi": 64, + "input_params": 43200, + "input_patch_count": 2, + "input_patch_params": 43200, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 64, + "left_input_sampled_relative_l2": 2.168208991570596e-6, + "n_gauss": 128, + "patch_build_secs": 0.028485206, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 9, + "raw_left_chi": 99, + "raw_left_params": 62240, + "raw_right_chi": 98, + "raw_right_params": 60416, + "rho_max": 8.0, + "right_chi": 64, + "right_input_sampled_relative_l2": 2.8526222945123154e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 0.3767062201175724 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.407904653, + "wall_times_secs": [ + 0.407904653 + ], + "max_error": 3.1360945365614698e-6, + "input_max_bond_dim": 64, + "output_max_bond_dim": 64, + "output_bond_dims": [ + 64 + ], + "n_params": 28028, + "n_patches": 1, + "max_patch_bond": 64, + "rtol": 1e-6, + "input_build_secs": 583.614199605, + "n_allocations": 209300, + "allocated_bytes": 363104096 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-global_fit-chi16.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-global_fit-chi16.json new file mode 100644 index 0000000..aa40daf --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-global_fit-chi16.json @@ -0,0 +1,64 @@ +{ + "schema_version": 1, + "case": "gaussian_mpo_contraction", + "algorithm": "global_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.10606601717798216, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n2-r6-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004611292, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.004500716, + "input_l2_rtol": 1e-6, + "input_params": 3740, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 16, + "left_input_max_patch_chi": 16, + "left_input_patch_count": 1, + "left_input_patch_params": 2008, + "left_input_sampled_relative_l2": 1.277785189967113e-6, + "n_gauss": 2, + "patch_build_secs": 0.011515116, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 6, + "raw_left_chi": 32, + "raw_left_params": 4640, + "raw_right_chi": 26, + "raw_right_params": 3872, + "rho_max": 8.0, + "right_chi": 16, + "right_input_max_patch_chi": 16, + "right_input_patch_count": 1, + "right_input_patch_params": 1732, + "right_input_sampled_relative_l2": 9.428700715139774e-7, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.007252303, + "wall_times_secs": [ + 0.007252303 + ], + "max_error": 5.94262639710347e-7, + "input_max_bond_dim": 16, + "output_max_bond_dim": 16, + "output_bond_dims": [ + 4, + 16, + 15, + 10, + 4 + ], + "n_params": 2008, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 37373, + "allocated_bytes": 12763689 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-global_fit-chi24.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-global_fit-chi24.json new file mode 100644 index 0000000..275c42f --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-global_fit-chi24.json @@ -0,0 +1,65 @@ +{ + "schema_version": 1, + "case": "gaussian_mpo_contraction", + "algorithm": "global_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.2121320343559643, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n8-r7-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004078569, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.002336353, + "input_l2_rtol": 1e-6, + "input_params": 7924, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 23, + "left_input_max_patch_chi": 23, + "left_input_patch_count": 1, + "left_input_patch_params": 3900, + "left_input_sampled_relative_l2": 1.7758182495895119e-6, + "n_gauss": 8, + "patch_build_secs": 0.016423873, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 7, + "raw_left_chi": 50, + "raw_left_params": 12720, + "raw_right_chi": 50, + "raw_right_params": 12720, + "rho_max": 8.0, + "right_chi": 24, + "right_input_max_patch_chi": 24, + "right_input_patch_count": 1, + "right_input_patch_params": 4024, + "right_input_sampled_relative_l2": 1.4847847441728126e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.014744834, + "wall_times_secs": [ + 0.014744834 + ], + "max_error": 7.31787434765934e-7, + "input_max_bond_dim": 24, + "output_max_bond_dim": 21, + "output_bond_dims": [ + 4, + 16, + 21, + 15, + 10, + 4 + ], + "n_params": 3652, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 42131, + "allocated_bytes": 30104620 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-global_fit-chi41.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-global_fit-chi41.json new file mode 100644 index 0000000..f626070 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-global_fit-chi41.json @@ -0,0 +1,66 @@ +{ + "schema_version": 1, + "case": "gaussian_mpo_contraction", + "algorithm": "global_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.4242640687119286, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n32-r8-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004661656, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.005393259, + "input_l2_rtol": 1e-6, + "input_params": 18380, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 41, + "left_input_max_patch_chi": 41, + "left_input_patch_count": 1, + "left_input_patch_params": 9272, + "left_input_sampled_relative_l2": 1.864016398611087e-6, + "n_gauss": 32, + "patch_build_secs": 0.025467873, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 8, + "raw_left_chi": 64, + "raw_left_params": 27592, + "raw_right_chi": 64, + "raw_right_params": 27592, + "rho_max": 8.0, + "right_chi": 40, + "right_input_max_patch_chi": 40, + "right_input_patch_count": 1, + "right_input_patch_params": 9108, + "right_input_sampled_relative_l2": 2.2322760961784926e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.040754897, + "wall_times_secs": [ + 0.040754897 + ], + "max_error": 1.3430384516297e-6, + "input_max_bond_dim": 41, + "output_max_bond_dim": 34, + "output_bond_dims": [ + 4, + 16, + 34, + 21, + 15, + 10, + 4 + ], + "n_params": 7340, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 47988, + "allocated_bytes": 64401888 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-global_fit-chi64.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-global_fit-chi64.json new file mode 100644 index 0000000..2eb9d64 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-global_fit-chi64.json @@ -0,0 +1,67 @@ +{ + "schema_version": 1, + "case": "gaussian_mpo_contraction", + "algorithm": "global_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.8485281374238572, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n128-r9-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.005568927, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.010572429, + "input_l2_rtol": 1e-6, + "input_params": 43200, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 64, + "left_input_max_patch_chi": 64, + "left_input_patch_count": 1, + "left_input_patch_params": 22184, + "left_input_sampled_relative_l2": 2.168208991570596e-6, + "n_gauss": 128, + "patch_build_secs": 0.044451815, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 9, + "raw_left_chi": 99, + "raw_left_params": 62240, + "raw_right_chi": 98, + "raw_right_params": 60416, + "rho_max": 8.0, + "right_chi": 64, + "right_input_max_patch_chi": 64, + "right_input_patch_count": 1, + "right_input_patch_params": 21016, + "right_input_sampled_relative_l2": 2.8526222945123154e-6, + "sigma_minor": 0.05, + "spacing": 3.0 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.202154557, + "wall_times_secs": [ + 0.202154557 + ], + "max_error": 1.5252138234932865e-6, + "input_max_bond_dim": 64, + "output_max_bond_dim": 59, + "output_bond_dims": [ + 4, + 16, + 59, + 36, + 21, + 15, + 10, + 4 + ], + "n_params": 17604, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 56339, + "allocated_bytes": 213042233 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-patched_fit-chi16.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-patched_fit-chi16.json new file mode 100644 index 0000000..35e4587 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-patched_fit-chi16.json @@ -0,0 +1,63 @@ +{ + "schema_version": 1, + "case": "gaussian_mpo_contraction", + "algorithm": "patched_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.10606601717798216, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n2-r6-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004611292, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.004500716, + "input_l2_rtol": 1e-6, + "input_params": 3740, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 16, + "left_input_max_patch_chi": 16, + "left_input_patch_count": 1, + "left_input_patch_params": 2008, + "left_input_sampled_relative_l2": 1.277785189967113e-6, + "n_gauss": 2, + "patch_build_secs": 0.011515116, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 6, + "raw_left_chi": 32, + "raw_left_params": 4640, + "raw_right_chi": 26, + "raw_right_params": 3872, + "rho_max": 8.0, + "right_chi": 16, + "right_input_max_patch_chi": 16, + "right_input_patch_count": 1, + "right_input_patch_params": 1732, + "right_input_sampled_relative_l2": 9.428700715139774e-7, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 1.1097758413790761 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.006534926, + "wall_times_secs": [ + 0.006534926 + ], + "max_error": 5.942626407271615e-7, + "input_max_bond_dim": 16, + "output_max_bond_dim": 16, + "output_bond_dims": [ + 16 + ], + "n_params": 2008, + "n_patches": 1, + "max_patch_bond": 16, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 37831, + "allocated_bytes": 11247267 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-patched_fit-chi24.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-patched_fit-chi24.json new file mode 100644 index 0000000..a181e19 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-patched_fit-chi24.json @@ -0,0 +1,63 @@ +{ + "schema_version": 1, + "case": "gaussian_mpo_contraction", + "algorithm": "patched_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.2121320343559643, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n8-r7-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004078569, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.002336353, + "input_l2_rtol": 1e-6, + "input_params": 7924, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 23, + "left_input_max_patch_chi": 23, + "left_input_patch_count": 1, + "left_input_patch_params": 3900, + "left_input_sampled_relative_l2": 1.7758182495895119e-6, + "n_gauss": 8, + "patch_build_secs": 0.016423873, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 7, + "raw_left_chi": 50, + "raw_left_params": 12720, + "raw_right_chi": 50, + "raw_right_params": 12720, + "rho_max": 8.0, + "right_chi": 24, + "right_input_max_patch_chi": 24, + "right_input_patch_count": 1, + "right_input_patch_params": 4024, + "right_input_sampled_relative_l2": 1.4847847441728126e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 0.9447692385567582 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.01560681, + "wall_times_secs": [ + 0.01560681 + ], + "max_error": 7.317874332117877e-7, + "input_max_bond_dim": 24, + "output_max_bond_dim": 21, + "output_bond_dims": [ + 21 + ], + "n_params": 3652, + "n_patches": 1, + "max_patch_bond": 21, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 47310, + "allocated_bytes": 25622537 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-patched_fit-chi41.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-patched_fit-chi41.json new file mode 100644 index 0000000..9ea073f --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-patched_fit-chi41.json @@ -0,0 +1,63 @@ +{ + "schema_version": 1, + "case": "gaussian_mpo_contraction", + "algorithm": "patched_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.4242640687119286, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n32-r8-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.004661656, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.005393259, + "input_l2_rtol": 1e-6, + "input_params": 18380, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 41, + "left_input_max_patch_chi": 41, + "left_input_patch_count": 1, + "left_input_patch_params": 9272, + "left_input_sampled_relative_l2": 1.864016398611087e-6, + "n_gauss": 32, + "patch_build_secs": 0.025467873, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 8, + "raw_left_chi": 64, + "raw_left_params": 27592, + "raw_right_chi": 64, + "raw_right_params": 27592, + "rho_max": 8.0, + "right_chi": 40, + "right_input_max_patch_chi": 40, + "right_input_patch_count": 1, + "right_input_patch_params": 9108, + "right_input_sampled_relative_l2": 2.2322760961784926e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 0.9656249674214774 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.04220572, + "wall_times_secs": [ + 0.04220572 + ], + "max_error": 1.773301387370064e-6, + "input_max_bond_dim": 41, + "output_max_bond_dim": 33, + "output_bond_dims": [ + 33 + ], + "n_params": 7192, + "n_patches": 1, + "max_patch_bond": 33, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 56270, + "allocated_bytes": 62181938 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-patched_fit-chi64.json b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-patched_fit-chi64.json new file mode 100644 index 0000000..cd6760f --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/raw/gaussian_mpo_contraction-patched_fit-chi64.json @@ -0,0 +1,63 @@ +{ + "schema_version": 1, + "case": "gaussian_mpo_contraction", + "algorithm": "patched_fit", + "params": { + "addition_tolerance": 1e-10, + "box_l": 0.8485281374238572, + "cache_hit": true, + "cache_key": "interpolative-gaussian-pair-v2-reva5157385f58048355a28a6a003d4cb63d3b1349b-n128-r9-sigma3fa999999999999a-rho4020000000000000-spacing4008000000000000-degree28-interp3ddb7cdfd9d7bdbb-add3ddb7cdfd9d7bdbb-seed0", + "cache_load_secs": 0.005568927, + "error_samples": 256, + "external_error_metric": "sampled_relative_l2", + "input_build_secs": 0.0, + "input_compression_secs": 0.010572429, + "input_l2_rtol": 1e-6, + "input_params": 43200, + "internal_tolerance": 1e-6, + "internal_tolerance_metric": "relative_l2_svd", + "interpolation_tolerance": 1e-10, + "left_chi": 64, + "left_input_max_patch_chi": 64, + "left_input_patch_count": 1, + "left_input_patch_params": 22184, + "left_input_sampled_relative_l2": 2.168208991570596e-6, + "n_gauss": 128, + "patch_build_secs": 0.044451815, + "patch_cap": 128, + "polynomial_degree": 28, + "r": 9, + "raw_left_chi": 99, + "raw_left_params": 62240, + "raw_right_chi": 98, + "raw_right_params": 60416, + "rho_max": 8.0, + "right_chi": 64, + "right_input_max_patch_chi": 64, + "right_input_patch_count": 1, + "right_input_patch_params": 21016, + "right_input_sampled_relative_l2": 2.8526222945123154e-6, + "sigma_minor": 0.05, + "spacing": 3.0, + "speedup_vs_global": 1.3811314491654347 + }, + "seed": 0, + "tolerance": 1e-6, + "wall_time_median_secs": 0.146368803, + "wall_times_secs": [ + 0.146368803 + ], + "max_error": 2.108072680627418e-6, + "input_max_bond_dim": 64, + "output_max_bond_dim": 58, + "output_bond_dims": [ + 58 + ], + "n_params": 16764, + "n_patches": 1, + "max_patch_bond": 58, + "rtol": 1e-6, + "input_build_secs": 0.0, + "n_allocations": 71594, + "allocated_bytes": 152143391 +} \ No newline at end of file diff --git a/result/linux-ryzen-ai-9-hx-470/report.md b/result/linux-ryzen-ai-9-hx-470/report.md new file mode 100644 index 0000000..77bd52b --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/report.md @@ -0,0 +1,70 @@ +# Tensor4all benchmark results + +Profile: `linux-ryzen-ai-9-hx-470`. CPU: AMD Ryzen AI 9 HX 470 w/ Radeon 890M. Threads: 1. CPU affinity: 0. Source revision: `05ebd78bdc40e09f1534a4bab2ac72131ce271d3`. tensor4all-rs revision: `a5157385f58048355a28a6a003d4cb63d3b1349b`. + +The source revision identifies the clean code that was measured. The commit adding these generated records necessarily follows that revision. + +All timings exclude input construction, cache I/O, format conversion, patch preparation, output conversion and accuracy evaluation. Gaussian inputs use independent two-dimensional interpolative decompositions, balanced pairwise addition, final relative-L2/SVD tolerance `1e-6`, and fixed patch cap 128. + +## Case 1: Fourier elementwise + +| input χ | K | arm | time (s) | sampled relative L2 | output χ | parameters | allocations | allocated (MiB) | +|---:|---:|---|---:|---:|---:|---:|---:|---:| +| 5 | 4 | aci | 0.051511 | 1.166e-08 | 9 | 1026 | 935093 | 92.31 | +| 5 | 4 | fit | 0.012445 | 1.213e-08 | 8 | 606 | 142866 | 23.60 | +| 5 | 4 | naive | 0.000918 | 1.211e-08 | 8 | 606 | 5323 | 2.61 | +| 5 | 4 | zipup | 0.006449 | 1.197e-08 | 8 | 606 | 80441 | 12.09 | +| 8 | 8 | aci | 0.073725 | 8.181e-09 | 16 | 1702 | 1293258 | 129.49 | +| 8 | 8 | fit | 0.012954 | 1.066e-08 | 8 | 726 | 143003 | 24.22 | +| 8 | 8 | naive | 0.001320 | 1.066e-08 | 8 | 726 | 5323 | 3.85 | +| 8 | 8 | zipup | 0.006676 | 1.080e-08 | 8 | 726 | 80573 | 12.56 | +| 8 | 16 | aci | 0.063717 | 6.086e-09 | 16 | 2206 | 1095665 | 111.28 | +| 8 | 16 | fit | 0.013237 | 1.620e-08 | 10 | 900 | 142989 | 24.91 | +| 8 | 16 | naive | 0.001710 | 1.620e-08 | 10 | 900 | 5323 | 5.12 | +| 8 | 16 | zipup | 0.006923 | 1.636e-08 | 10 | 900 | 80547 | 13.03 | +| 11 | 32 | aci | 0.074269 | 4.931e-09 | 23 | 3318 | 1253379 | 136.21 | +| 11 | 32 | fit | 0.013954 | 1.806e-08 | 13 | 1200 | 143389 | 26.29 | +| 11 | 32 | naive | 0.002830 | 1.806e-08 | 13 | 1200 | 5346 | 8.20 | +| 11 | 32 | zipup | 0.007409 | 2.001e-08 | 13 | 1182 | 80911 | 14.02 | +| 14 | 64 | aci | 0.070777 | 4.480e-09 | 29 | 4752 | 1149586 | 129.83 | +| 14 | 64 | fit | 0.014982 | 1.813e-08 | 16 | 1638 | 143592 | 28.05 | +| 14 | 64 | naive | 0.004957 | 1.813e-08 | 16 | 1638 | 5363 | 13.72 | +| 14 | 64 | zipup | 0.007748 | 1.856e-08 | 16 | 1638 | 81012 | 15.25 | +| 16 | 128 | aci | 0.071915 | 3.247e-09 | 32 | 6808 | 1115945 | 135.96 | +| 16 | 128 | fit | 0.022185 | 1.044e-08 | 18 | 2258 | 143793 | 31.91 | +| 16 | 128 | naive | 0.011175 | 1.044e-08 | 18 | 2258 | 5374 | 23.45 | +| 16 | 128 | zipup | 0.010941 | 1.064e-08 | 18 | 2258 | 81754 | 18.20 | + +## Case 2: Gaussian elementwise, global versus patched + +| input χ | raw χ | N | R | arm | time (s) | sampled relative L2 | input patches | input max patch χ | output patches | parameters | allocations | allocated (MiB) | speedup | +|---:|---:|---:|---:|---|---:|---:|---:|---:|---:|---:|---:|---:|---:| +| 16 | 32 | 2 | 6 | global_aci | 0.004388 | 1.398e-06 | 2 | 16 | 1 | 6560 | 57157 | 11.46 | 1.000 | +| 16 | 32 | 2 | 6 | global_fit | 0.007957 | 2.293e-06 | 2 | 16 | 1 | 2300 | 57856 | 13.03 | 1.000 | +| 16 | 32 | 2 | 6 | patched_aci | 0.007052 | 2.292e-06 | 2 | 16 | 1 | 2300 | 82738 | 17.63 | 0.622 | +| 16 | 32 | 2 | 6 | patched_fit | 0.013680 | 2.292e-06 | 2 | 16 | 1 | 2300 | 82983 | 24.03 | 0.582 | +| 24 | 50 | 8 | 7 | global_aci | 0.011431 | 2.942e-06 | 2 | 24 | 1 | 23268 | 90766 | 29.69 | 1.000 | +| 24 | 50 | 8 | 7 | global_fit | 0.013576 | 3.462e-06 | 2 | 24 | 1 | 4776 | 65431 | 22.37 | 1.000 | +| 24 | 50 | 8 | 7 | patched_aci | 0.014890 | 3.459e-06 | 2 | 24 | 1 | 4908 | 117462 | 39.44 | 0.768 | +| 24 | 50 | 8 | 7 | patched_fit | 0.033404 | 3.459e-06 | 2 | 24 | 1 | 4908 | 110716 | 55.35 | 0.406 | +| 41 | 64 | 32 | 8 | global_aci | 0.034806 | 3.468e-06 | 2 | 41 | 1 | 98992 | 107105 | 75.53 | 1.000 | +| 41 | 64 | 32 | 8 | global_fit | 0.040618 | 4.661e-06 | 2 | 41 | 1 | 11072 | 99690 | 56.84 | 1.000 | +| 41 | 64 | 32 | 8 | patched_aci | 0.028499 | 4.366e-06 | 2 | 41 | 1 | 11740 | 111749 | 68.33 | 1.221 | +| 41 | 64 | 32 | 8 | patched_fit | 0.106389 | 4.365e-06 | 2 | 41 | 1 | 11740 | 157617 | 149.48 | 0.382 | +| 64 | 99 | 128 | 9 | global_aci | 0.149108 | 2.705e-06 | 2 | 64 | 1 | 317712 | 139715 | 226.03 | 1.000 | +| 64 | 99 | 128 | 9 | global_fit | 0.153660 | 3.993e-06 | 2 | 64 | 1 | 25476 | 110346 | 174.30 | 1.000 | +| 64 | 99 | 128 | 9 | patched_aci | 0.069497 | 3.179e-06 | 2 | 64 | 1 | 28028 | 140133 | 134.67 | 2.146 | +| 64 | 99 | 128 | 9 | patched_fit | 0.407905 | 3.136e-06 | 2 | 64 | 1 | 28028 | 209300 | 346.28 | 0.377 | + +## Case 3: Gaussian MPO-MPO contraction, global versus patched + +| input χ | raw χ | N | R | arm | time (s) | sampled relative L2 | input patches | input max patch χ | output patches | parameters | allocations | allocated (MiB) | speedup | +|---:|---:|---:|---:|---|---:|---:|---:|---:|---:|---:|---:|---:|---:| +| 16 | 32 | 2 | 6 | global_fit | 0.007252 | 5.943e-07 | 2 | 16 | 1 | 2008 | 37373 | 12.17 | 1.000 | +| 16 | 32 | 2 | 6 | patched_fit | 0.006535 | 5.943e-07 | 2 | 16 | 1 | 2008 | 37831 | 10.73 | 1.110 | +| 24 | 50 | 8 | 7 | global_fit | 0.014745 | 7.318e-07 | 2 | 24 | 1 | 3652 | 42131 | 28.71 | 1.000 | +| 24 | 50 | 8 | 7 | patched_fit | 0.015607 | 7.318e-07 | 2 | 24 | 1 | 3652 | 47310 | 24.44 | 0.945 | +| 41 | 64 | 32 | 8 | global_fit | 0.040755 | 1.343e-06 | 2 | 41 | 1 | 7340 | 47988 | 61.42 | 1.000 | +| 41 | 64 | 32 | 8 | patched_fit | 0.042206 | 1.773e-06 | 2 | 41 | 1 | 7192 | 56270 | 59.30 | 0.966 | +| 64 | 99 | 128 | 9 | global_fit | 0.202155 | 1.525e-06 | 2 | 64 | 1 | 17604 | 56339 | 203.17 | 1.000 | +| 64 | 99 | 128 | 9 | patched_fit | 0.146369 | 2.108e-06 | 2 | 64 | 1 | 16764 | 71594 | 145.10 | 1.381 | diff --git a/result/linux-ryzen-ai-9-hx-470/run.yaml b/result/linux-ryzen-ai-9-hx-470/run.yaml new file mode 100644 index 0000000..f63eec6 --- /dev/null +++ b/result/linux-ryzen-ai-9-hx-470/run.yaml @@ -0,0 +1,9 @@ +profile: linux-ryzen-ai-9-hx-470 +date_utc: 2026-10-10T22:45:59Z +os: Linux +chip: AMD Ryzen AI 9 HX 470 w/ Radeon 890M +memory_gb: 58.5 +repo_rev: 05ebd78bdc40e09f1534a4bab2ac72131ce271d3 +tensor4all_rs_rev: a5157385f58048355a28a6a003d4cb63d3b1349b +threads: 1 +cpu_affinity: 0