Metraipolis › archive › dc_95ec2ae048da

Beyond atol and rtol: Sound Test Oracles for Low-Precision GPU Kernels — arXiv:2610.38983

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.38983
License (as archived)
CC0-1.0
Archived
2026-11-01T02:29:07.853Z
SHA-256 of the archived text
475d59d3e2a636e04711d30d91c6ec4b75a4d06b73019be81468b410348dab3e
Stable URL of this page
https://metraipolis.com/documento/dc_95ec2ae048da

How to cite

Beyond atol and rtol: Sound Test Oracles for Low-Precision GPU Kernels — arXiv:2610.38983 — arXiv (en), https://arxiv.org/abs/2610.38983 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_95ec2ae048da (Metraipolis archive dc_95ec2ae048da, sha256 475d59d3e2a636e04711d30d91c6ec4b75a4d06b73019be81468b410348dab3e)
BibTeX
@misc{dc95ec2ae048da,
  title = {Beyond atol and rtol: Sound Test Oracles for Low-Precision GPU Kernels — arXiv:2610.38983},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.38983},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_95ec2ae048da; sha256 475d59d3e2a636e04711d30d91c6ec4b75a4d06b73019be81468b410348dab3e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.38983. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Low-precision GPU kernels are commonly tested against a reference using hand-picked absolute and relative tolerances. Legitimate numerical error varies with the input, format and precision plan, so a fixed tolerance trades false alarms for missed faults. We present numoracle, which derives per-element acceptance regions from precision-annotated reference programs. Its calculus tracks shared rounding errors, provenance-ordered products, block-scaled quantizers and calibrated math-library errors. A rounding-interval check detects faulty stores that symmetric bounds can admit. Soundness is conditional on explicit rounding, hardware, domain and implementation assumptions; a fixed-input mode removes the zero-mean rounding assumption. We evaluate 13 Triton operators from Liger-Kernel, Unsloth and vLLM on two GPU generations. The main campaigns produce no false alarm on 33,960 correct-kernel and behavior-preserving-variant runs per GPU. Under input shift, the GPU campaigns record no false alarm while tuned tolerances flag up to 30% of correct runs; final-version reference checks identify inputs without complete finite-range certificates. On the RTX PRO 6000, numoracle kills 116 of 124 effective mutants, versus 59-83 for four fixed baselines. Among 400 tuned tolerance configurations, requiring at most 1% false alarms in every evaluated Liger input group permits at most 31 of 53 main-mutant kills, versus numoracle's 50. Historical replays expose accuracy faults in eight change pairs without rejecting their matched corrected controls; a ninth change concerns rounding parity.

Tema degli esploratori: calcolo e infrastrutture (elenco arXiv cs.SE, nuovi lavori del giorno).
Categorie arXiv: Mathematical Software (cs.MS) ; Programming Languages (cs.PL); Software Engineering (cs.SE).
Pagina del lavoro: https://arxiv.org/abs/2610.38983 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-01T02:12:58.508Z dalla pagina https://arxiv.org/list/cs.SE/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).