Metraipolis › archive › dc_7f596d356aee

PEAT: Pseudo-Error Assessment for GPU Kernel Validation in DNN Training — arXiv:2609.13544

Source

arXiv (en)

Original

https://arxiv.org/abs/2609.13544

License (as archived)

CC0-1.0

Archived

2026-10-15T23:17:32.557Z

SHA-256 of the archived text

1f884cad95f5303570ac4e889a271d2630963b55af9d807b20643212c1dbad74

Stable URL of this page

https://metraipolis.com/documento/dc_7f596d356aee

How to cite

PEAT: Pseudo-Error Assessment for GPU Kernel Validation in DNN Training — arXiv:2609.13544 — arXiv (en), https://arxiv.org/abs/2609.13544 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7f596d356aee (Metraipolis archive dc_7f596d356aee, sha256 1f884cad95f5303570ac4e889a271d2630963b55af9d807b20643212c1dbad74)

@misc{dc7f596d356aee,
  title = {PEAT: Pseudo-Error Assessment for GPU Kernel Validation in DNN Training — arXiv:2609.13544},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.13544},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7f596d356aee; sha256 1f884cad95f5303570ac4e889a271d2630963b55af9d807b20643212c1dbad74},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.13544. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Deep neural networks (DNNs) are widely adopted in various fields, driving an emerging trend in developing software stacks associated with DNN training systems. For example, many codes have been ported across different frameworks or developed to leverage the computing power of GPUs or domain-specific accelerators. However, validating a kernel implementation in DNN training is time-consuming and generally requires massive storage. Specifically, this poses a fundamental question: how to characterize the behavior of a new implementation when it is integrated into a DNN training flow. Unfortunately, this problem is not well investigated in the literature, to the best of our knowledge. To address this shortcoming, we present PEAT - a lightweight inspection framework for \underline{P}seudo-\underline{E}rror \underline{A}ssessment associated with GPU kernel validation in DNN \underline{T}raining. Firstly, inspired by conventional fault injection (FI), PEAT's Profiler invokes an operation-wise kernel in a training flow to collect a DNN model's states (e.g., checkpoints and activations). More importantly, the Profiler introduces two simple yet effective techniques, playback FI and frequency-based runtime FI, leveraging persistent kernel calling during the training process. Secondly, PEAT's Analyzer characterizes profiled errors, revealing some signatures from the error distribution of a kernel compared to the golden one. Lastly, PEAT's Detector provides some guidelines as a sufficient condition, which enables associating several well-known error models with signature patterns. We demonstrate the applicability of our approach by presenting the results and analysis using GPUs from the two most popular vendors, NVIDIA V100 and AMD MI250, on various AI models, from vision tasks to language models, for both pretraining and finetuning scenarios.

Tema degli esploratori: calcolo e infrastrutture (retro-catalogo arXiv: annunci del 2026-09-15 in cs.DC).
Categorie arXiv: Distributed, Parallel, and Cluster Computing (cs.DC) ; Computational Engineering, Finance, and Science (cs.CE); Software Engineering (cs.SE).
Pagina del lavoro: https://arxiv.org/abs/2609.13544 — 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 2026-10-15T22:54:12.950Z dalla pagina https://arxiv.org/catchup/cs.DC/2026-09-15?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).