Metraipolis › archive › dc_33a883eb5710
Architecting for Hardware Diversity: Meta's End-to-End Co-Design for High-Performance AI Training on AMD GPUs — arXiv:2610.22230
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.22230
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T10:04:39.112Z
- SHA-256 of the archived text
c20b30e2fc3c470ce62457129a67ed2411e0d611fc37c2c351a49892547a4382- Stable URL of this page
https://metraipolis.com/documento/dc_33a883eb5710
How to cite
Architecting for Hardware Diversity: Meta's End-to-End Co-Design for High-Performance AI Training on AMD GPUs — arXiv:2610.22230 — arXiv (en), https://arxiv.org/abs/2610.22230 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_33a883eb5710 (Metraipolis archive dc_33a883eb5710, sha256 c20b30e2fc3c470ce62457129a67ed2411e0d611fc37c2c351a49892547a4382)
BibTeX
@misc{dc33a883eb5710,
title = {Architecting for Hardware Diversity: Meta's End-to-End Co-Design for High-Performance AI Training on AMD GPUs — arXiv:2610.22230},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.22230},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_33a883eb5710; sha256 c20b30e2fc3c470ce62457129a67ed2411e0d611fc37c2c351a49892547a4382},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.22230. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Meta deploys a heterogeneous fleet comprising NVIDIA GPUs, AMD GPUs, and proprietary MTIA accelerators to reduce costs and mitigate vendor lock-in. However, the extensive history of NVIDIA-specific optimizations hinders performance portability to other hardware. This paper presents our methods for optimizing alternative accelerators, using AMD Instinct MI350X versus NVIDIA B200 for recommendation model training as a case study. We propose three reusable principles. First, hardware-model co-design is essential: rather than directly porting models from NVIDIA GPUs, revising model structures and software recipes allows full exploitation of accelerator-specific strengths (e.g., MI350X's 1.6x HBM capacity relative to B200). Second, our analytical performance model enables zero-warmup optimal kernel selection across diverse datatypes and matrix shapes in production models, drastically simplifying tuning. Third, increasingly mature agentic workflows have lowered the barrier to porting from the CUDA ecosystem; we extend these workflows for microarchitecture-aware kernel optimizations and custom shim generation. Compared to the directly ported baseline, these methods deliver up to 1.5x improvement in end-to-end training throughput. To our knowledge, this is the first hyperscaler report on productionizing AMD GPUs for large-scale training. Our contributions have been upstreamed to the open-source PyTorch, Triton, and ROCm ecosystems. Tema degli esploratori: calcolo e infrastrutture (elenco arXiv cs.DC, nuovi lavori del giorno). Categorie arXiv: Distributed, Parallel, and Cluster Computing (cs.DC) ; Hardware Architecture (cs.AR); Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2610.22230 — 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 2026-10-16T08:45:02.687Z dalla pagina https://arxiv.org/list/cs.DC/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).