Performance Optimization of Baryon-block Construction in the Stochastic LapH Method

TL;DR

This paper presents optimized implementation strategies for baryon-block computation in the stochastic LapH method, achieving a 7.2x speedup to enhance efficiency in lattice QCD spectroscopy calculations.

Contribution

The paper introduces new implementation strategies that significantly accelerate baryon-block computations in the stochastic LapH method.

Findings

Achieved a 7.2x speedup over previous implementation.

Demonstrated effective utilization of modern hardware capabilities.

Improved computational efficiency for lattice QCD spectroscopy.

Abstract

Implementations of measurement kernels in high-level Lattice QCD frameworks enable rapid prototyping, but can leave hardware capabilities significantly underutilized. This is an acceptable tradeoff if the time spent in unoptimized routines is generally small. The computational cost of modern spectroscopy projects however can be comparable to or even exceed the cost of generating gauge configurations and computing solutions of the Dirac equation. One such key kernel in the stochastic LapH method is the computation of baryon blocks; we discuss several implementation strategies and achieve a 7.2x speedup over the current implementation on a system with Intel(R) Xeon(R) Platinum 8358 processors, formerly Ice Lake.