High-performance computing (HPC) systems require massive amounts of power, and inefficient code can lead to excessive energy usage. By improving code efficiency, less energy is required to perform computations, leading to reduced environmental impact and lower energy costs. HPC systems consume substantial resources, including electricity and cooling infrastructure. These resources are not evenly distributed globally, and their availability can be limited in certain regions. Just as large scale computing systems can be used more efficiently with optimized code, efficiency practices can improve power consumption for all parts of our power infrastructure. This approach promotes a more sustainable use of computing resources, aligning with the principles of energy justice and reducing the environmental impact of HPC systems.
In this project, participants will learn how to optimize energy consumption in computing systems. Participants will delve into programming languages, specifically Python, to optimize code for parallel processing, enabling efficient utilization of system resources. Additionally, participants will explore energy efficiency in our power infrastructure and how it is measured and improved.
Project Leaders and Trainers:
Charles Lively
Dan Fulton
Lipi Gupta
Rollin Thomas
Yun (Helen) He
Image source: https://cacm.acm.org/magazines/2010/3/76284-toward-energy-efficient-computing/fulltext