Performance of a Lattice Quantum Chromodynamics Kernel on the Cell Processor

TL;DR

This paper demonstrates a Lattice Quantum Chromodynamics kernel optimized for the Cell processor, achieving significant computational performance and highlighting its potential for future scientific calculations.

Contribution

It provides a detailed implementation and performance analysis of a Lattice QCD kernel on the Cell processor, showcasing its suitability for high-performance scientific computing.

Findings

Achieved up to 45 GFlop/s per socket performance

Identified key issues in porting Lattice QCD to the Cell processor

Indicated the Cell processor's potential for future Lattice QCD calculations

Abstract

The implementation of a proof-of-concept Lattice Quantum Chromodynamics kernel on the Cell processor is described in detail, illustrating issues encountered in the porting process. The resulting code performs up to 45GFlop/s per socket, indicating that the Cell processor is likely to be a good platform for future Lattice QCD calculations.