Aurora Exascale Supercomputer Uses Intel® Advisor Roofline Feature
Aurora at Argonne National Laboratory is one of the US Department of Energy’s exascale supercomputers that will be deployed in 2022. oneAPI provides all essential components for porting applications to Aurora with optimal performance.
The roofline features in Intel® Advisor (part of the Intel® oneAPI Base Toolkit) provide intuitive performance analysis results on Intel® GPUs and useful insights about performance bottlenecks for further optimization. This session presents Intel Advisor use cases from workloads that include molecular dynamics and computational fluid dynamics.
JaeHyuk Kwack works in the performance engineering group at the Argonne Leadership Computing Facility. He received his Bachelor of Science and Master of Science degrees in engineering from Seoul National University, South Korea. He earned his PhD and completed his postdoctoral training in computational mechanics for computational fluid dynamics and fluid solid interaction problems at the University of Illinois at Urbana-Champaign. At Argonne, he has been focused on the OpenMP* offloading model and associated performance tools and math libraries for the upcoming Aurora system. Before joining Argonne in 2018, he worked for the Blue Waters supercomputing project at the National Center for Supercomputing Applications.
Zakhar A. Matveev, PhD, is a principal engineer and product architect for the Intel Advisor tool on x86 CPUs and Intel GPUs. His focus and professional interests are in the areas of high-performance computing, parallel programming, hardware and software co-design, and computer graphics.