Performance Modeling, Benchmarking and Simulation of High Performance Computer Systems

11:30 - 12:00
Evaluating the Performance of NVIDIA’s A100 Ampere GPU for Sparse and Batched Computations

Hartwig Anzt, Yuhsiang M. Tsai, Terry Cojean
Karlsruhe Institute of Technology, Germany

Ahmad Abdelfattah
University of Tennessee, TN

Jack Dongarra
University of Tennessee, TN
Oak Ridge National Laboratory, TN
University of Manchester, United Kingdom

GPU accelerators have become an important backbone for scientific high performance-computing, and the performance advances obtained from adopting new GPU hardware are significant. In this paper we take a first look at NVIDIA’s newest server-line GPU, the A100 architecture, part of the Ampere generation. Specifically, we assess its performance for sparse and batch computations, as these routines are relied upon in many scientific applications, and compare to the performance achieved on NVIDIA’s previous server-line GPU.