Affiliation:
1. Mathematical Institute, University of Oxford, Oxford, UK
2. Oxford e-Research Centre, University of Oxford, Oxford, UK
Abstract
High-performance computing has evolved remarkably over the past 20 years, and that progress is likely to continue. However, in recent years, this progress has been achieved through greatly increased hardware complexity with the rise of multicore and manycore processors, and this is affecting the ability of application developers to achieve the full potential of these systems. This article outlines the key developments on the hardware side, both in the recent past and in the near future, with a focus on two key issues: energy efficiency and the cost of moving data. It then discusses the much slower evolution of system software, and the implications of all of this for application developers.
Subject
General Physics and Astronomy,General Engineering,General Mathematics
Reference36 articles.
1. Top500. 2014 Top 500 supercomputer sites. See http://www.top500.org/.
2. Heroux M& Dongarra J. Toward a new metric for ranking high performance computing systems. Technical report SAND2013-4744. Albuquerque NM: Sandia National Laboratories. See http://www.sandia.gov/maherou/docs/HPCG-Benchmark.pdf.
3. Advanced Scientific Computing Research (ASCR) Co-Design See http://science.energy.gov/ascr/research/scidac/co-design/ascr/research/scidac/co-design/ (accessed 4 February 2014).
4. IBM Power Systems S812L and S822L See http://www-03.ibm.com/systems/power/hardware/s812l-s822l/specs.html (accessed 14 May 2014).
5. Kestor G Gioiosa R Kerbyson D& Hoisie A. 2013 Quantifying the energy cost of data movement in scientific applications. 56–65. (doi:10.1109/IISWC.2013.6704670).
Cited by
31 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献