PCon Preparing applications for El Capitan and beyond
As Lawrence Livermore National Laboratories (LLNL) prepares to stand up its next supercomputer, El Capitan, application teams prepare to pivot to another GPU architecture.
This talk presents how the LLNL application teams made the transition from distributed-memory, CPU-only architectures to GPUs. They share institutional best practices. They discuss new open-source software products as tools for porting and profiling applications and as avenues for collaboration across the computational science community.
Join LLNL's Erik Draeger and Jane Herriman, who presented this talk at Pawsey's PaCER Conference in September 2023.
Licence: Creative Commons Attribution 4.0
Keywords: GPUs, supercomputing, HPC, PaCER
Target audience: Masters student, PhD student, Academic, Post-doc / Fellow, Professional (research-related), Professional (other), Undergraduate
Date modified: 2023-09-22
Date published: 2023-09-29
Contributors: Pawsey Supercomputing Research Centre