Abstract
Scientific applications do frequently suffer from limited compute performance. In this article, we investigate the suitability of specialized computer chips to overcome this limitation. An enhanced Poisson Boltzmann program is ported to the graphics processing unit and the application specific integrated circuit MDGRAPE-3 and resulting execution times are compared to the conventional performance obtained on a modern central processing unit. Speed Up factors are measured and an analysis of numerical accuracy is provided. On both specialized architectures the improvement is increasing with problem size and reaches up to a Speed Up factor of 39 x for the largest problem studied. This type of alternative high performance computing can significantly improve the performance of demanding scientific applications.
| Original language | English |
|---|---|
| Pages (from-to) | 2351-2357 |
| Number of pages | 7 |
| Journal | Journal of Computational Chemistry |
| Volume | 30 |
| Issue number | 14 |
| DOIs | |
| Publication status | Published - 2009 Nov 15 |
Fingerprint
Keywords
- ASIC
- GPU
- Implicit solvation models
- MDGRAPE-3
- Poisson boltzmann
- Solvation
ASJC Scopus subject areas
- Chemistry(all)
- Computational Mathematics
Cite this
Current performance gains from utilizing the GPU or the ASIC MDGRAPE-3 within an enhanced poisson boltzmann approach. / Narumi, Tetsu; Yasuoka, Kenji; Taiji, Makoto; Höfinger, Siegfried.
In: Journal of Computational Chemistry, Vol. 30, No. 14, 15.11.2009, p. 2351-2357.Research output: Contribution to journal › Article
}
TY - JOUR
T1 - Current performance gains from utilizing the GPU or the ASIC MDGRAPE-3 within an enhanced poisson boltzmann approach
AU - Narumi, Tetsu
AU - Yasuoka, Kenji
AU - Taiji, Makoto
AU - Höfinger, Siegfried
PY - 2009/11/15
Y1 - 2009/11/15
N2 - Scientific applications do frequently suffer from limited compute performance. In this article, we investigate the suitability of specialized computer chips to overcome this limitation. An enhanced Poisson Boltzmann program is ported to the graphics processing unit and the application specific integrated circuit MDGRAPE-3 and resulting execution times are compared to the conventional performance obtained on a modern central processing unit. Speed Up factors are measured and an analysis of numerical accuracy is provided. On both specialized architectures the improvement is increasing with problem size and reaches up to a Speed Up factor of 39 x for the largest problem studied. This type of alternative high performance computing can significantly improve the performance of demanding scientific applications.
AB - Scientific applications do frequently suffer from limited compute performance. In this article, we investigate the suitability of specialized computer chips to overcome this limitation. An enhanced Poisson Boltzmann program is ported to the graphics processing unit and the application specific integrated circuit MDGRAPE-3 and resulting execution times are compared to the conventional performance obtained on a modern central processing unit. Speed Up factors are measured and an analysis of numerical accuracy is provided. On both specialized architectures the improvement is increasing with problem size and reaches up to a Speed Up factor of 39 x for the largest problem studied. This type of alternative high performance computing can significantly improve the performance of demanding scientific applications.
KW - ASIC
KW - GPU
KW - Implicit solvation models
KW - MDGRAPE-3
KW - Poisson boltzmann
KW - Solvation
UR - http://www.scopus.com/inward/record.url?scp=70349459442&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=70349459442&partnerID=8YFLogxK
U2 - 10.1002/jcc.21257
DO - 10.1002/jcc.21257
M3 - Article
VL - 30
SP - 2351
EP - 2357
JO - Journal of Computational Chemistry
JF - Journal of Computational Chemistry
SN - 0192-8651
IS - 14
ER -