A Massive Data Parallel Computational Framework on Petascale/Exascale Hybrid Computer Systems
Applications Department, Poznan Supercomputing and Networking Center, Poland
ParCo 2011
@article{blazewicz2011massive,
title={A Massive Data Parallel Computational Framework on Petascale/Exascale Hybrid Computer Systems},
author={BLAZEWICZ, M. and BRANDT, S.R. and DIENER, P. and KOPPELMAN, D.M. and KUROWSKI, K. and L{"O}FFLER, F. and SCHNETTER, E. and TAO, J.},
year={2011}
}
Heterogeneous systems are becoming more common on High Performance Computing (HPC) systems. Even using tools like CUDA [1] and OpenCL [2] it is a non-trivial task to obtain optimal performance on the GPU. Approaches to simplifying this task include Merge [3] (a library based framework for heterogeneous multi-core systems), Zippy [4] (a framework for parallel execution of codes on multiple GPU’s), BSGP [5] (a new programming language for general purpose computation on the GPU) and CUDA-lite [6] (an enhancement to CUDA that transforms code based on annotations). In addition, efforts are underway to improve compiler tools for automatic parallelization and optimization of affine loop nests for GPU’s [7] and for automatic translation of OpenMP parallelized codes to CUDA [8]. In this paper we present an alternative approach: a new computational framework for the development of massively data parallel scientific codes applications suitable for use on such petascale/exascale hybrid systems built upon the highly scalable Cactus framework [9,10] As the first non-trivial demonstration of its usefulness, we successfully developed a new 3D CFD code that achieves improved performance.
October 4, 2011 by hgpu
