Integrate reviewed feature 01d004c385.
Single official TBB pair: 4134.657782 to 3747.788261 seconds (9.356748% lower).
Numeric bits and per-increment Newton counts match; no new regression.
Existing 16 FAIL / 1 SKIP, two missing references and physical-validation
limitations remain. Full validation and independent review are preserved in
.build/jet-product-20261021-01/. No remote push.
Record adopted outer-product, accumulation and scratch changes and the
nonadoption reasons for scalar/channel/arena candidates.
Report the fresh TBB single-pair result: 4134.657782 to 3747.788261 seconds,
9.356748 percent lower elapsed, with unchanged numeric bits and Newton counts.
Link raw evidence and retain 16 baseline failures, one skip, two missing
formal references, physical-validation gaps and existing performance targets.
Date-scope previous MKL/HHT evidence instead of replacing its history.
Documentation only. Product source bytes match the fully verified 685f5e6;
git diff --check and static report/link review passed. No build/test rerun.
Record T4 as a nonadopted request-specific derivative-channel candidate.
No product, private helper, public API, fixture or test-registration changes
are required; the production sources and archives remain identical to T3.
Original-source preservation tests show that Compose with a high nonfinite
third coefficient and zero direction retains finite value/gradient/Hessian
but produces mixed-second NaN, IsFinite failure and INVALID FE/MXCSR state.
An overflowing high reciprocal coefficient reaches the same boundary.
Negative scalar times zero also preserves signed-zero directional/mixed
channels, so replacing unused fields with initialized +0 changes behavior.
The omission candidate was rejected without a validation-replication framework.
Copied-source instrumentation in three valid representative small HHT tests
recorded API entry attempts before validation: none 42, explicit-zero 78,
nonzero 47. These are scoped test-entry observations, not successful or
accepted evaluation counts or official-model frequency estimates. Normal
Release request-mode timings are descriptive auxiliary costs, not channel
cost attribution or a before/after speedup. No official model ran in T4.
Separate T4 original/current supplemental comparisons passed in six routes;
six explicit omission-mutant comparisons returned the expected exit 1.
Unchanged T3 build/default/related evidence is linked by exact source/archive
hashes. Existing physics failures and full T6 verification remain unchanged.
Artifact root: C:/git/FESADev/.build/jet-product-20261021-01/t4
T4-summary.json SHA256:
c348a62d6c686c667286ba8ee16f310134b13c779132fe839765def89af568c0
T4-summary.md SHA256:
ecfea9771dd8de396d86394262f393a3da9c480b8a80c500a9545a383b80bb15
commands-manifest.json SHA256:
4cace21f3b49932f52d9cf1bd9dfdc2ad7d9ccf40657da8849c678dd24e0a538
사용자 후속 결정에 따라 이번 실험을 변위 검증 통과와 실패10개 보류로 수용한다.
기존 전체 검사 FAIL/SKIP과 수치 기준을 보존한다.
전체101프레임 q/v/a/A/energy 차이와 관측 시간·outer 횟수를 추가한다.
공식 성능 조건·전체 비용·시간 정확도 미확보는 후속 과제로 남긴다.
문서만 변경: source450/reference34 hash 불변, 로컬 링크87개와 정적 검사 통과.
T7 executed Native Debug 883/72/43 tests and full CMake Debug CTest998;
both retain 10 failures and 1 resolution skip. Python145 passed.
All five original Hex8 Release accuracy CLI/comparator pairs returned0.
Source574 and formal reference34 inventories/SHA remain unchanged.
No tracked file diff; operational evidence is retained under
.build/tolerance-experiment/go-20261018/T7/verification.json.
This evidence commit does not claim numerical suitability, complete R8,
official candidate speedup, or missing-reference validation.
KOKO\Mimi