A Case Study in Coordination Programming: Performance Evaluation of S-Net vs Intel's Concurrent Collections

TL;DR

This paper compares the performance of S-Net and Intel's CnC coordination languages using tiled Cholesky decomposition, showing S-Net's superior performance on a 48-core system.

Contribution

It provides a detailed performance evaluation of S-Net versus CnC for a complex linear algebra problem, highlighting the advantages of S-Net's separation of concerns.

Findings

S-Net outperforms CnC on a 48-core machine.

Two S-Net implementations are described and evaluated.

Cnc implementations with and without tuning are compared.

Abstract

We present a programming methodology and runtime performance case study comparing the declarative data flow coordination language S-Net with Intel's Concurrent Collections (CnC). As a coordination language S-Net achieves a near-complete separation of concerns between sequential software components implemented in a separate algorithmic language and their parallel orchestration in an asynchronous data flow streaming network. We investigate the merits of S-Net and CnC with the help of a relevant and non-trivial linear algebra problem: tiled Cholesky decomposition. We describe two alternative S-Net implementations of tiled Cholesky factorization and compare them with two CnC implementations, one with explicit performance tuning and one without, that have previously been used to illustrate Intel CnC. Our experiments on a 48-core machine demonstrate that S-Net manages to outperform CnC on this problem.