State-of-the-art in Heterogeneous Computing

André R. Brodtkorb, Christopher Dyken, Trond Runar Hagen, Jon M. Hjelmervik, Olaf O. Storaasli

DOAJ (DOAJ: Directory of Open Access Journals) · 2010 · 27 citations · 0 references

Concepts

TL;DR

Node‑level heterogeneous architectures are attractive for their high performance and energy or cost efficiency, and the rise of fine‑grained parallelism in HPC and workstations has created an urgent need for a comprehensive overview. The paper aims to survey the state‑of‑the‑art in heterogeneous computing, focusing on Cell, GPU, and FPGA architectures, and to compare them qualitatively and quantitatively. The authors review hardware, software tools, and state‑of‑the‑art techniques and algorithms for Cell, GPU, and FPGA architectures. They provide a qualitative and quantitative comparison of the architectures and discuss the future direction of heterogeneous computing.

Abstract

Node level heterogeneous architectures have become attractive during the last decade for several reasons: compared to traditional symmetric CPUs, they offer high peak performance and are energy and/or cost efficient. With the increase of fine-grained parallelism in high-performance computing, as well as the introduction of parallelism in workstations, there is an acute need for a good overview and understanding of these architectures. We give an overview of the state-of-the-art in heterogeneous computing, focusing on three commonly found architectures: the Cell Broadband Engine Architecture, graphics processing units (GPUs), and field programmable gate arrays (FPGAs). We present a review of hardware, available software tools, and an overview of state-of-the-art techniques and algorithms. Furthermore, we present a qualitative and quantitative comparison of the architectures, and give our view on the future of heterogeneous computing.