Wilson and Domainwall Kernels on Oakforest-PACS

TL;DR

This paper evaluates the performance of Wilson and Domainwall kernels on the Oakforest-PACS supercomputer, highlighting implementation strategies and comparing various approaches including the Grid library and Bridge++ integration.

Contribution

It provides performance analysis of Wilson and Domainwall kernels on a new high-performance Intel Xeon Phi system with diverse implementation methods.

Findings

Performance benchmarks of kernels on Oakforest-PACS

Comparison of implementation strategies including Grid and Bridge++

Insights into the efficiency of kernels on Intel Xeon Phi architecture

Abstract

We report the performance of Wilson and Domainwall Kernels on a new Intel Xeon Phi Knights Landing based machine named Oakforest-PACS, which is co-hosted by University of Tokyo and Tsukuba University and is currently fastest in Japan. This machine uses Intel Omni-Path for the internode network. We compare performance with several types of implementation including that makes use of the Grid library. The code is incorporated with the code set Bridge++.