Metraipolis › archive › dc_4dfd126a6bdd

The Joule Point: an Energy-Optimal Operating Point for AI Inference — arXiv:2609.27926

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.27926
License (as archived): CC0-1.0
Archived: 2026-10-08T10:07:28.584Z
SHA-256 of the archived text: 153860eef6f46fbbfc80c125da91b057d1a86d4748708eba3669ad38ad8bf023
Stable URL of this page: https://metraipolis.com/documento/dc_4dfd126a6bdd

How to cite

The Joule Point: an Energy-Optimal Operating Point for AI Inference — arXiv:2609.27926 — arXiv (en), https://arxiv.org/abs/2609.27926 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_4dfd126a6bdd (Metraipolis archive dc_4dfd126a6bdd, sha256 153860eef6f46fbbfc80c125da91b057d1a86d4748708eba3669ad38ad8bf023)

BibTeX

@misc{dc4dfd126a6bdd,
  title = {The Joule Point: an Energy-Optimal Operating Point for AI Inference — arXiv:2609.27926},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.27926},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_4dfd126a6bdd; sha256 153860eef6f46fbbfc80c125da91b057d1a86d4748708eba3669ad38ad8bf023},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.27926. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Data centers serving AI inference strive to maximize GPU utilization, running their cards at full power by default: this maximizes throughput and holds latencies down, but it is energy-inefficient, spending more energy per inference than the same work needs at a better operating point. That inefficiency is a scheduling choice, not a hardware limit. The cause is physical: for a given workload, a GPU's board power rises superlinearly along its power-performance curve, on top of a fixed power floor that is paid for as long as the job runs, so energy per inference is U-shaped in the operating point (GPU, power cap). The minimum, which we name the Joule Point, is a cap at 43 to 46 per cent of a large GPU's rated power; capping to it cuts energy per inference by roughly a quarter to a third at a modest cost in capital and latency: each request runs about 1.2 times slower, and holding aggregate throughput takes that same factor more cards. Under load, the Joule Point is nearly a per-card-type constant, so a single static cap per card type captures nearly all the saving (mean penalty under one per cent), turning the online per-job search that prior systems run into a one-time characterization. We ground this in ELF, a dense power-cap dataset that sweeps 20 inference models across four GPUs; replaying it in a fleet simulation under a power budget, capping each job to the least power meeting its deadline spends 18 to 45 per cent less energy per served job. These results recast data-center energy as a schedulable resource an operator can adapt to a budget, an electricity price, or a carbon signal while meeting its service targets.

Tema degli esploratori: calcolo e infrastrutture (retro-catalogo arXiv: annunci del 2026-09-24 in cs.PF).
Categorie arXiv: Performance (cs.PF).
Pagina del lavoro: https://arxiv.org/abs/2609.27926 — 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 2026-10-08T09:14:54.774Z dalla pagina https://arxiv.org/catchup/cs.PF/2026-09-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).