Tuan Truong Vu, Cheng Tan · 2026 · arXiv
DOI: 10.48550/arXiv.2610.28429 · open-access PDF
Los núcleos de GPU son críticos para el rendimiento de los servicios de IA de hoy en día, pero los mismos operadores se reescriben repetidamente: para cada generación de GPU, formato numérico, régimen de servicio y lenguaje de núcleo. Cada reescritura debe preservar la función calculada por la implementación que sustituye. Hoy en día, esta obligación depende casi totalmente de la prueba diferencial: ejecute el nuevo núcleo y una referencia confiable en tensores mostrados aleatoriamente, luego compare sus salidas dentro de una tolerancia.
Machine translation (Meta M2M100); the English original follows.
GPU kernels are critical to the performance of today's AI services, yet the same operators are rewritten repeatedly: for each GPU generation, numeric format, serving regime, and kernel language. Every rewrite must preserve the function computed by the implementation it replaces. Today, this obligation depends almost entirely on differential testing: run the new kernel and a trusted reference on randomly sampled tensors, then compare their outputs within a tolerance. This approach has blind spots that additional test cases cannot eliminate. For example, a logical bug may produce an error smalle
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.28429, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange