The Community for Technology Leaders
Green Image
Issue No. 08 - August (2012 vol. 45)
ISSN: 0018-9162
pp: 26-32
Wen-mei W. Hwu , University of Illinois at Urbana-Champaign
Geng Liu , University of Illinois at Urbana-Champaign
Nasser Anssari , University of Illinois at Urbana-Champaign
I-Jui Sung , University of Illinois at Urbana-Champaign
Li-Wen Chang , University of Illinois at Urbana-Champaign
Christopher Rodrigues , University of Illinois at Urbana-Champaign
John A. Stratton , University of Illinois at Urbana-Champaign
Nady Obeid , KLA-Tencor
ABSTRACT
A study of the implementation patterns among massively threaded applications for many-core GPUs reveals that each of the seven most commonly used algorithm and data optimization techniques can enhance the performance of applicable kernels by 2 to 10× in current processors while also improving future scalability. The featured Web extra is a video interview with author John Stratton, who describes how implementation patterns can improve future scalability. YouTube URL: http://youtu.be/fgn9LJbInMw
INDEX TERMS
Instruction sets, System-on-a-chip, Bandwidth, Histograms, Optimization, Graphics processing unit, Multithreading, Parboil benchmarks, massively threaded systems, optimization patterns, accelerators, scalability
CITATION
Wen-mei W. Hwu, Geng Liu, Nasser Anssari, I-Jui Sung, Li-Wen Chang, Christopher Rodrigues, John A. Stratton, Nady Obeid, "Algorithm and Data Optimization Techniques for Scaling to Massively Threaded Systems", Computer, vol. 45, no. , pp. 26-32, August 2012, doi:10.1109/MC.2012.194
82 ms
(Ver )