Presentation
SIGN IN TO VIEW THIS PRESENTATION Sign In
Exploring GPU-to-GPU Communication: Insights into Supercomputer Interconnects
DescriptionMulti-GPU nodes are increasingly common in the rapidly evolving landscape of exascale supercomputers. On these systems, GPUs on the same node are connected through dedicated networks, with bandwidths up to a few terabits per second. However, gauging performance expectations and maximizing system efficiency is challenging due to different technologies, design options, and software layers. This paper comprehensively characterizes three supercomputers — Alps, Leonardo, and LUMI — each with a unique architecture and design. We focus on performance evaluation of intra-node and inter-node interconnects on up to 4096 GPUs, using a mix of intra-node and inter-node benchmarks. By analyzing its limitations and opportunities, we aim to offer practical guidance to researchers, system architects, and software developers dealing with multi-GPU supercomputing. Our results show that there is untapped bandwidth, and there are still many opportunities for optimization, ranging from network to software optimization.
Authors
Event Type
Paper
TimeTuesday, 19 November 20244:30pm - 5pm EST
LocationB312-B313A
Accelerators
HPC Infrastructure
Performance Evaluation and/or Optimization Tools
State of the Practice
TP
Archive
view
Similar Presentations