Sustainable Computing for HPC and AI
HPCWire, Monday, September 23rd, 2024
Reducing energy consumption and associated CO2 emissions has become a priority for organizations of all sizes. Climate concerns are now top-of-mind among customers, employees, and shareholders, and as a result, most organizations have made public commitments to reduce emissions as part of corporate ESG programs.
Driven by new online services, crypto-currencies, and rapid advances in AI, energy consumption from data centers is on the rise, outpacing our ability to bring renewable sources online in many parts of the world. According to the International Energy Agency (IEA), global data center energy consumption is expected to double, growing from an estimated 460 terawatt-hours (TWh) in 2022 to more than 1,000 TWh by 2026.[i]
This article explains how advanced scheduling techniques pioneered in HPC can help organizations boost efficiency and reduce energy requirements for their simulation and AI workloads.