GPU Optimization of Large-Scale Eigenvalue Solver

Pavel Kůs,Hermann Lederer,Andreas Marek

Lecture Notes in Computational Science and Engineering（2019）

Max Planck Computing and Data Facility

被引用8|浏览19

摘要

We present a GPU implementation of a large-scale eigenvalue solver as a part of the ELPA library. We describe the methodology of utilizing the GPU accelerators within an already well optimized MPI-based code. We present numerical results using two different HPC systems equipped with modern GPU accelerators and show the performance benefits of the GPU version.

更多

查看译文

关键词

GPU Computing,Performance Optimization,Eigenvalue Problems,High-Performance Computing