Metraipolis › archive › dc_2bf2136a694d

One Simple Trick for Improving the Performance of Energy-Limited Local Inference and Training — arXiv:2609.11936

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.11936
License (as archived): CC0-1.0
Archived: 2026-10-16T03:08:38.874Z
SHA-256 of the archived text: d1f6e0353c177c3f4fa5d05cdfcccf9747f4675b9c68d06803d87c7f12e3cabd
Stable URL of this page: https://metraipolis.com/documento/dc_2bf2136a694d

How to cite

One Simple Trick for Improving the Performance of Energy-Limited Local Inference and Training — arXiv:2609.11936 — arXiv (en), https://arxiv.org/abs/2609.11936 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2bf2136a694d (Metraipolis archive dc_2bf2136a694d, sha256 d1f6e0353c177c3f4fa5d05cdfcccf9747f4675b9c68d06803d87c7f12e3cabd)

BibTeX

@misc{dc2bf2136a694d,
  title = {One Simple Trick for Improving the Performance of Energy-Limited Local Inference and Training — arXiv:2609.11936},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.11936},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2bf2136a694d; sha256 d1f6e0353c177c3f4fa5d05cdfcccf9747f4675b9c68d06803d87c7f12e3cabd},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.11936. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Energy supply and heat dissipation are two of the main challenges with modern GPU deployments. While typically discussed in the context of new datacenter constructions, the same constraints also apply to small form-factor consumer devices, such as the DGX spark. In workloads characterized by alternating compute-intensive tasks such as matmuls with memory-bound operations such as norms or cross-entropy, the compute-intensive parts might hit power and/or thermal limits and start throttling. In this short paper, we show that chunking the workload into smaller parts that alternate compute and memory in higher frequencies, these power and temperature spikes can be smoothed out, preventing throttling and resulting in considerably faster wall-clock time and reduced total energy consumption. We present several scenarios in which this effect can be exploited on a DGX Spark with up to 2% performance and energy improvements, and demonstrate that the same phenomenon also happens on less constrained systems, such as a multi-GPU server, albeit at significantly reduced effect size of 1-2%.

Tema degli esploratori: calcolo e infrastrutture (retro-catalogo arXiv: annunci del 2026-09-14 in cs.DC).
Categorie arXiv: Performance (cs.PF) ; Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.11936 — 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 2026-10-16T03:03:05.094Z dalla pagina https://arxiv.org/catchup/cs.DC/2026-09-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).