Scheduling Variable Capacity Resources for Sustainability Workshop - Inria - Institut national de recherche en sciences et technologies du numérique
Rapport Année : 2023

Scheduling Variable Capacity Resources for Sustainability Workshop

Résumé

We gathered the research community focused on compute scheduling in the new age of renewable power generation -- where variation in weather and solar radiation drives renewable generation, and thereby the carbon-intensity of power, and efforts to reduce environmental damage then create variation in compute capacity (opportunity). In this age, effective utilization of that opportunity is the key to sustainable computing. We report important findings on the attributes of future platforms, characterizing the power grid variation, reliability challenges, and two-way relationships between datacenters and grid that are emerging. Workloads are changing as well, with growing understanding of how computing can be malleable, delay flexible, or even acceptably approximate. Between these, we focus on how scheduling can meet these challenges, managing between flexible workloads and dynamically varying platforms – performing to new metrics that reflect both performance and non-performance attributes (e.g. carbon emissions). The concerns that underlie a shift to variable capacity derive from a broader societal concern about the sustainability of computing, and our last section addresses critical challenges in awareness, responsibilities, and action. The talks and position papers from attendees are available at https://people.cs.uchicago.edu/~aachien/workshops/varsched23/.
Fichier principal
Vignette du fichier
Varsched23-Final-Report-20230709.pdf (372.24 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04159509 , version 1 (11-07-2023)

Licence

Identifiants

  • HAL Id : hal-04159509 , version 1

Citer

Anne Benoit, Andrew A Chien, Yves Robert. Scheduling Variable Capacity Resources for Sustainability Workshop. ROMA (INRIA Rhône-Alpes / LIP Laboratoire de l’Informatique du Parallélisme); University of Chicago. 2023. ⟨hal-04159509⟩
129 Consultations
143 Téléchargements

Partager

More