Metraipolis › archive › dc_7c997b60d1bc

A principled approach for energy-efficient training via phase-aware GPU frequency tuning — arXiv:2609.24205

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.24205
License (as archived): CC0-1.0
Archived: 2026-10-09T06:53:05.890Z
SHA-256 of the archived text: ef545323ddb25184ad7063c343e4d78d91207d957537cfb54b2303d7bc22791f
Stable URL of this page: https://metraipolis.com/documento/dc_7c997b60d1bc

How to cite

A principled approach for energy-efficient training via phase-aware GPU frequency tuning — arXiv:2609.24205 — arXiv (en), https://arxiv.org/abs/2609.24205 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7c997b60d1bc (Metraipolis archive dc_7c997b60d1bc, sha256 ef545323ddb25184ad7063c343e4d78d91207d957537cfb54b2303d7bc22791f)

BibTeX

@misc{dc7c997b60d1bc,
  title = {A principled approach for energy-efficient training via phase-aware GPU frequency tuning — arXiv:2609.24205},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.24205},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7c997b60d1bc; sha256 ef545323ddb25184ad7063c343e4d78d91207d957537cfb54b2303d7bc22791f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.24205. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Modern AI model training imposes unprecedented computational demands, making it a key contributor to datacenter energy consumption. Yet a significant fraction of the energy consumed during training does not translate to useful computation due to bottlenecks throughout the training pipeline. We present PAFT, a phase-aware, dynamically adaptable GPU frequency tuning system that reduces energy consumption of training workloads with minimal performance overhead. The key insight behind PAFT is that bottlenecks represent an energy optimization opportunity, rather than purely a performance problem: when GPUs are bound to stall, PAFT opportunistically reduces their clock frequencies to match the pace of bottlenecked devices, saving energy without impacting execution time. PAFT achieves this by continuously monitoring pipeline behavior and applying fine-grained frequency adjustments, adapting to workload and system changes. Experiments conducted on twelve widely used models show that PAFT consistently outperforms all baselines, achieving energy savings of up to 46% with an average overhead of 4%.

Tema degli esploratori: calcolo e infrastrutture (retro-catalogo arXiv: annunci del 2026-09-22 in cs.DC).
Categorie arXiv: Distributed, Parallel, and Cluster Computing (cs.DC) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.24205 — 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 2026-10-09T06:50:13.541Z dalla pagina https://arxiv.org/catchup/cs.DC/2026-09-22?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).