Exécution de notre outil de synthèse sur un nouvel article
Vérifiez l’article ci-dessous, puis lancez sa synthèse.
Characterizing Power Management Opportunities for LLMs in the Cloud
Pratyush Patel (Microsoft (United States), University of Washington), Esha Choukse (Microsoft (United States)), Chaojie Zhang (Microsoft (United States)), Íñigo Goiri (Microsoft (United States)), Brijesh Warrier (Microsoft (United States)), Nithish Mahalingam (Microsoft (United States)), Ricardo Bianchini (Microsoft (United States))
Résumé : Recent innovation in large language models (LLMs), and their myriad use cases have rapidly driven up the compute demand for datacenter GPUs. Several cloud providers and other enterprises plan to substantially grow their datacenter capacity to support these new workloads. A key bottleneck resource in datacenters is power, which LLMs are quickly saturating due to their rapidly increasing model sizes.