***Note: Venue changed to Pick Hall 022!***
Join us for a Lunch & Learn with Computer Science PhD student Varsha Rao.
Lunch is provided.
Abstract
EasyC: Modeling the Carbon Footprint of HPC Systems
The carbon footprint of computing systems is growing rapidly, accelerated by the increasing use of generative AI. However, it is difficult to even perform carbon assessment of a single system using Green House Gas (GHG) Protocol carbon emission accounting methodology, and effectively infeasible for a collection of systems. As a result, there is little to no carbon reporting for High Performance Computing (HPC) systems, and even the largest HPC sites do not do GHG Protocol reporting.
We assess the carbon footprint of HPC, focusing on the Top 500 systems. The key challenge lies in modeling the carbon footprint with limited data availability.
We propose EasyC, a novel carbon accounting framework and tool that requires only a few key data metrics. With the disclosed Top500.org data, and using EasyC, we were able to model the operational carbon of 391 HPC systems and the embodied carbon of 283 HPC systems. We further show how this coverage can be enhanced by additional public information. With improved coverage, interpolation is used to produce the first carbon footprint estimates of the Top 500 HPC systems. They are 1.4 million MT CO2e operational carbon (1 Year) and 1.9 million MT CO2e embodied carbon. We also project how the Top 500’s carbon footprint will increase through 2030. Further, we explore the data availability and enhancements, showing that coverage can be increased to 98% of Top 500 systems for operational carbon and 80.8% for embodied carbon emissions.