Data-parallel programming with Intel Array Building Blocks (ArBB)

TL;DR

This paper introduces Intel Array Building Blocks (ArBB), a high-level data-parallel programming environment, demonstrating its scalability and portability through benchmarks on multi-core platforms compared to OpenMP and MKL.

Contribution

It presents the porting of key scientific kernels to ArBB and evaluates their performance and scalability on a multi-core system.

Findings

ArBB achieves scalable performance on multi-core systems.

Performance comparisons show ArBB's competitiveness with OpenMP and MKL.

Demonstrates portability of scientific kernels across platforms.

Abstract

Intel Array Building Blocks is a high-level data-parallel programming environment designed to produce scalable and portable results on existing and upcoming multi- and many-core platforms. We have chosen several mathematical kernels - a dense matrix-matrix multiplication, a sparse matrix-vector multiplication, a 1-D complex FFT and a conjugate gradients solver - as synthetic benchmarks and representatives of scientific codes and ported them to ArBB. This whitepaper describes the ArBB ports and presents performance and scaling measurements on the Westmere-EX based system SuperMIG at LRZ in comparison with OpenMP and MKL.