{ "machine": "x86_64", "records": [ { "workers": 3, "block": 768, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.001184228, "cpu_seconds": 0.001586, "peak_observations": 3, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "baseline", "repeat": 0 }, { "workers": 3, "block": 768, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.000780818, "cpu_seconds": 0.001407, "peak_observations": 3, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "baseline", "repeat": 1 }, { "workers": 3, "block": 768, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.000735996, "cpu_seconds": 0.001454, "peak_observations": 3, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "baseline", "repeat": 2 }, { "workers": 3, "block": 1, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.004854976, "cpu_seconds": 0.013673, "peak_observations": 2, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "candidate", "repeat": 0 }, { "workers": 3, "block": 1, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.003973085, "cpu_seconds": 0.011009, "peak_observations": 2, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "candidate", "repeat": 1 }, { "workers": 3, "block": 1, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.005139563, "cpu_seconds": 0.014419, "peak_observations": 1, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "candidate", "repeat": 2 }, { "workers": 3, "block": 4, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.001665869, "cpu_seconds": 0.004219, "peak_observations": 2, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "candidate", "repeat": 0 }, { "workers": 3, "block": 4, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.001695678, "cpu_seconds": 0.004347, "peak_observations": 3, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "candidate", "repeat": 1 }, { "workers": 3, "block": 4, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.001606282, "cpu_seconds": 0.00412, "peak_observations": 2, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "candidate", "repeat": 2 }, { "workers": 3, "block": 16, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.001115221, "cpu_seconds": 0.002522, "peak_observations": 3, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "candidate", "repeat": 0 }, { "workers": 3, "block": 16, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.001223608, "cpu_seconds": 0.002794, "peak_observations": 3, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "candidate", "repeat": 1 }, { "workers": 3, "block": 16, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.001242774, "cpu_seconds": 0.002861, "peak_observations": 3, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "candidate", "repeat": 2 }, { "workers": 3, "block": 128, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.000680106, "cpu_seconds": 0.001334, "peak_observations": 3, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "candidate", "repeat": 0 }, { "workers": 3, "block": 128, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.000800628, "cpu_seconds": 0.001663, "peak_observations": 3, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "candidate", "repeat": 1 }, { "workers": 3, "block": 128, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.000814975, "cpu_seconds": 0.001622, "peak_observations": 2, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "candidate", "repeat": 2 }, { "workers": 3, "block": 768, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.000851628, "cpu_seconds": 0.001724, "peak_observations": 3, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "candidate", "repeat": 0 }, { "workers": 3, "block": 768, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.000749958, "cpu_seconds": 0.001412, "peak_observations": 3, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "candidate", "repeat": 1 }, { "workers": 3, "block": 768, "replicates": 4, "samples": 768, "iterations": 0, "wall_seconds": 0.000553179, "cpu_seconds": 0.001152, "peak_observations": 3, "mean": 0.9984095168134586, "variance": 1.225201421274993e-05, "kind": "candidate", "repeat": 2 }, { "workers": 3, "block": 768, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.505244901, "cpu_seconds": 1.027335, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "baseline", "repeat": 0 }, { "workers": 3, "block": 768, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.506098546, "cpu_seconds": 1.015039, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "baseline", "repeat": 1 }, { "workers": 3, "block": 768, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.50698729, "cpu_seconds": 1.02912, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "baseline", "repeat": 2 }, { "workers": 3, "block": 1, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.345829247, "cpu_seconds": 1.034111, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "candidate", "repeat": 0 }, { "workers": 3, "block": 1, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.342609994, "cpu_seconds": 1.024877, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "candidate", "repeat": 1 }, { "workers": 3, "block": 1, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.343030362, "cpu_seconds": 1.022994, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "candidate", "repeat": 2 }, { "workers": 3, "block": 4, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.34047087, "cpu_seconds": 1.017868, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "candidate", "repeat": 0 }, { "workers": 3, "block": 4, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.344676822, "cpu_seconds": 1.030746, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "candidate", "repeat": 1 }, { "workers": 3, "block": 4, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.34027383, "cpu_seconds": 1.016232, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "candidate", "repeat": 2 }, { "workers": 3, "block": 16, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.342691653, "cpu_seconds": 1.019627, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "candidate", "repeat": 0 }, { "workers": 3, "block": 16, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.347872659, "cpu_seconds": 1.020672, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "candidate", "repeat": 1 }, { "workers": 3, "block": 16, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.347116053, "cpu_seconds": 1.024895, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "candidate", "repeat": 2 }, { "workers": 3, "block": 128, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.343258336, "cpu_seconds": 1.020959, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "candidate", "repeat": 0 }, { "workers": 3, "block": 128, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.344691745, "cpu_seconds": 1.027037, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "candidate", "repeat": 1 }, { "workers": 3, "block": 128, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.344221553, "cpu_seconds": 1.025586, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "candidate", "repeat": 2 }, { "workers": 3, "block": 768, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.506588548, "cpu_seconds": 1.013659, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "candidate", "repeat": 0 }, { "workers": 3, "block": 768, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.506331555, "cpu_seconds": 1.017238, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "candidate", "repeat": 1 }, { "workers": 3, "block": 768, "replicates": 4, "samples": 768, "iterations": 20000, "wall_seconds": 0.507962133, "cpu_seconds": 1.027488, "peak_observations": 3, "mean": 3.7953040747795854, "variance": 1.2713872783845792e-05, "kind": "candidate", "repeat": 2 }, { "workers": 3, "block": 16, "replicates": 1, "samples": 768, "iterations": 20000, "wall_seconds": 0.253747653, "cpu_seconds": 0.253987, "peak_observations": 1, "mean": 3.7917856490150528, "variance": 0, "kind": "baseline", "repeat": 0 }, { "workers": 3, "block": 16, "replicates": 1, "samples": 768, "iterations": 20000, "wall_seconds": 0.095650739, "cpu_seconds": 0.276964, "peak_observations": 3, "mean": 3.7917856490150528, "variance": 0, "kind": "candidate", "repeat": 0 } ], "date": "2026-11-01", "baseline_revision": "1a2153e24b2062acb45dc869d399f951c6c0922f", "uni20_revision": "1cb2474eb0a54e13ec7d5f2c1823752020b2425c", "cpu": "Intel Core i7-12700KF", "compiler": "GCC 13.3.0, -O3 -DNDEBUG -std=c++23", "precision": "binary64", "scope": "Synthetic CPU scheduling benchmark; not a production six-spinon speedup claim. One compiler process also ran on this 20-logical-CPU host. No affinity pinning. Three repeats per four-replicate case.", "kernel_source": "tests/benchmark_monte_carlo.cpp", "kernel_source_sha256": "8dbf8fc21aacc34d84d0eedf369b3ac524e77ab755113ee896daaa7ca69ce93b", "admission_window": "2 * workers", "source_comparison": "baseline uses the four original scheduling/sampling headers from the baseline revision; candidate uses sample blocks. Identical compile flags and libraries.", "precision_replay_check": { "formats": [ "binary64", "native long double", "MPFR160" ], "blocks": [ 1, 4, 16, 128 ], "scope": "Independent baseline/candidate programs: complete and numerical-failure results, means, covariance, first failures, counts, maxima, metadata. No timing fields.", "identical_output_sha256": "66b1439ac760454b12deb4193a87bc7261df4c150b44083ee78552c45a237de9" }, "physical_check": { "arguments": [ "--sectors", "6", "--momentum", "1.57079632679489661923132169163975", "--omega", "2", "--precision", "bits-384", "--six-cutoff", "8", "--replicates", "4", "--samples", "32", "--seed", "42", "--density-relative-tolerance", "1e-6", "--threads", "3", "--diagnostics", "--replicate-output", "--format", "json" ], "binary_sha256": "624615aad68c159438b6e8112498951d4601303a70577ee9d90c9553428f7823", "scope": "One small finite-R=8 physical query, two runs per block size in reversed order. Same candidate binary for all blocks. Block=32 is one task per replicate. Spectrum and all replicate rows match exactly. No production R=1040 speedup claim. The executable was compiled directly against candidate headers with immutable installed-runtime-build dependencies; its embedded Bethe revision identifies that reused generated header, not this candidate.", "records": [ { "repeat": 0, "block": 32, "process_wall_seconds": 11.469138633983675, "query_wall_seconds": 11.187203480000008, "query_cpu_seconds": 26.831457, "status": "finite_domain_estimate", "failed_points": 0, "export_sha256": "40cdc006307e45bb7e9cf5f6f20b4b0425cbc92fb985749a6ea4f817c1617fb7" }, { "repeat": 0, "block": 16, "process_wall_seconds": 9.904770939989248, "query_wall_seconds": 9.652326344000002, "query_cpu_seconds": 26.755104, "status": "finite_domain_estimate", "failed_points": 0, "export_sha256": "a6f1331825990398cb05ab44db5f8f0502e3b8a41cf0dca9a64ed7936b1642dd" }, { "repeat": 0, "block": 4, "process_wall_seconds": 13.781125774985412, "query_wall_seconds": 13.530287720000004, "query_cpu_seconds": 26.500074, "status": "finite_domain_estimate", "failed_points": 0, "export_sha256": "7c3ce5a9e3f26004bea82a5ba37e34c7264b1b532dd157f6f8957eddbc0c734a" }, { "repeat": 0, "block": 1, "process_wall_seconds": 21.481774079002207, "query_wall_seconds": 21.230450352000005, "query_cpu_seconds": 26.413671, "status": "finite_domain_estimate", "failed_points": 0, "export_sha256": "65458e2072fe2e46696a7d0ff4c614224dd0da1ed8d9865b901c7cad4fe4b20c" }, { "repeat": 1, "block": 1, "process_wall_seconds": 21.431559382006526, "query_wall_seconds": 21.17668874399999, "query_cpu_seconds": 26.484314, "status": "finite_domain_estimate", "failed_points": 0, "export_sha256": "411f8d75bac95d4eeefc8239beb4aa09da981769024f196a6fdee7f95b725fdb" }, { "repeat": 1, "block": 4, "process_wall_seconds": 14.00534056298784, "query_wall_seconds": 13.740720095, "query_cpu_seconds": 26.848519, "status": "finite_domain_estimate", "failed_points": 0, "export_sha256": "b11101933b1d8783dee0daa4b061793eb287a46a0136de9273553781dd828232" }, { "repeat": 1, "block": 16, "process_wall_seconds": 9.441834509023465, "query_wall_seconds": 9.187876030000012, "query_cpu_seconds": 26.868983, "status": "finite_domain_estimate", "failed_points": 0, "export_sha256": "7a0f28f5a116e0e0ff820491058bd9a48b2b8ae7561f98a867b3ef3bb58c6dff" }, { "repeat": 1, "block": 32, "process_wall_seconds": 11.391490287991473, "query_wall_seconds": 11.134854286000007, "query_cpu_seconds": 26.76743, "status": "finite_domain_estimate", "failed_points": 0, "export_sha256": "bd12e6dcc2107c4fa47a1843f31c75953756e6d11018ad25e9efd2f61f31240c" } ] }, "validation": { "revision": "bbfdd612841d0cdf0165d25f8f29d587dd7d1bc2", "scope": "Initial validation at bbfdd61, before the terminal-replicate and cancellation regressions were added. The follow-up sections below record later 15- and 17-test suites; historical counts and measurements are retained.", "unit_tests_passed": 74, "cli_suites": [ "check_xxx_spinon_sectors.py", "check_xxx_spinon_moments.py" ], "sanitizers": "14 Monte Carlo tests with GCC AddressSanitizer and UndefinedBehaviorSanitizer, leak detection enabled. Dependency libraries reused without instrumentation.", "replay_sha256": "66b1439ac760454b12deb4193a87bc7261df4c150b44083ee78552c45a237de9" }, "header_sha256": { "monte_carlo.hpp": "58ce239c6846298cf6e3e214108233f3a98d179419b7f7c919532e3bcac026c0", "observable_grid.hpp": "628f814fdf4ab2a7c4efe94e207f0f61cc3b28a8643f46e5bf0b45cb9e2dbd84", "sampling.hpp": "4f434f72979d9945cb6ff08ff548a709e341d1ac18a1ff0c8dfa379531e54abf", "execution.hpp": "30ada51e7d1b778d43e7f80f968d0355bc3686296b91385975db9d8d4b7673d8" }, "candidate_revision": "bbfdd612841d0cdf0165d25f8f29d587dd7d1bc2", "terminal_replicate_followup": { "review": "https://github.com/Uni20-dev/bethe/pull/90#discussion_r4433409339", "scope": "Newly admitted blocks skip observation work after collection records a terminal replicate. Already-admitted blocks drain; other replicates continue. Original performance measurements above remain measurements of bbfdd61.", "regression": "TerminalReplicatesDoNotStartNewObservationBlocks", "baseline": "bbfdd61 fails: 17 integrand calls for the failed replicate with window=1 and window=3.", "fixed": "One call with window=1; no more calls than the admitted window with window=3; both healthy replicates complete all 65 samples.", "tests": "All 15 Monte Carlo tests pass, including ASan/UBSan with leak detection; independent baseline/candidate oracle still matches exactly for fp64/long-double/MPFR160 and grains 1/4/16/128.", "header_sha256": "d59d849a7548503a96bfe635f0cd23de743bf9e21e6712555a896ad1657f585b", "logs": { "terminal-block-baseline.log": "68a5c20a379708b6224144f7c4a213006543951288c3b3d28480e20be3f4767c", "monte-carlo-terminal-fix.log": "0aa8e4e5edc30ab9fda40ffebdbff33a6667b99f2a325efdfc2347df6d284d07", "monte-carlo-terminal-fix-san.log": "7d691346eb9d35de42d365294084d168a09e47824fd9f1cee5c845ae2cddd29b" }, "revision": "e2e2e93b1744a706072ab7b688c7deb97e99f689" }, "cancellation_followup": { "review": "https://github.com/Uni20-dev/bethe/pull/90#discussion_r4433711166", "base_revision": "e2e2e93b1744a706072ab7b688c7deb97e99f689", "scope": "After draining a cancelled grid, mark unfinished rows cancelled when no earlier failure exists. Complete rows and unadmitted replicates are unchanged. No new physical timing measurement.", "regression": "CancellationAtBlockBoundaryClassifiesUnfinishedReplicate fails on the base at one and two workers: completed=4, cancelled=false, failed_sample absent. The fixed row records cancelled=true, failed_sample=4 and no means. Complete four-sample rows keep their means and the unadmitted second replicate stays absent.", "count_semantics": "CancelledCountsIncludeCompletedSpeculativeBlocks forces sample zero to enter before later blocks request a stop. Completed count=9 includes sample zero and two later four-sample blocks; first missing index=1 remains separate. No incomplete-row mean, overall estimate or covariance is exposed.", "tests": "All 17 Monte Carlo tests pass normally and with GCC AddressSanitizer/UndefinedBehaviorSanitizer and leak detection; both cancellation cases pass 50 repeats in each build. Reused dependency libraries are not sanitizer-instrumented.", "fixture_history": "The initial speculative-work fixture did not ensure sample zero entered before the stop. Repetition exposed that test race; the final fixture explicitly synchronizes entry. Original logs are retained locally.", "header_sha256": "b14cd49b975fd2ce463092c22e7b5c1891a1ce835537939b0372ec5ef0a5325c", "test_source_sha256": "0f978219dfe6216c5b58f021976565708faf5269ef49f5bc973c3f8616ff7097", "logs": { "before2-regression.log": "9c05e63ec0d5e1a8486923ea99b01a36bc20dc0a5ef8f1d99a105133f32e4b2e", "after2-tests.log": "ad4339b7b6101ea0d1d35dd516bf4f450be3f4a5ee691362d59a9a8a8add5374", "after2-cancellation-repeat.log": "fec65760963dc3844f7e6262cfe4d40bbc10a1b00e644ae05f01ac06852cf8e8", "san-final-tests.log": "b2b3574fb83bf42339576c1627cd797b48ce05d6c6dd98d42b13314ae035231f", "san-final-cancellation-repeat.log": "5963ec666bf18e8fa9fd42cc24135c6b47d43360cc827d58631ba25fd1b05da7" }, "revision": "616e539bf7b0ba1b41ae5dc841a83495a1d2fd9f" }, "benchmark_header_followup": { "review": "https://github.com/Uni20-dev/bethe/pull/90#discussion_r4433805601", "change": "Explicit standard-library includes in the benchmark, including ; numerical implementation unchanged. Historical benchmark measurements and kernel hash above remain pinned to their original revision.", "benchmark_source_sha256": "c576be17417abf51e3a0dd99fcb995b6bfce6b182d193a9e89689ef76dd68aa8", "validation": "GCC 13 benchmark rebuilt; four correctness smokes at three workers and four replicates, block sizes 768/16 and iterations 0/20000 exactly match the original means and variances for the same workload. Timings from these smokes are not substituted for the original performance measurements.", "smoke_record_sha256": "aec25c8f4ef6b30483dc18bd06d9612c21092e37da1e7559984197b41a92169e" } }