The Case of the Missing Supercomputer Performance: Achieving Optimal Performance on the 8,192 Processors of ASCI Q Export

@inproceedings{citeulike:3393568, abstract = {In this paper we describe how we improved the effective performance of ASCI Q, the world's second-fastest supercomputer, to meet our expectations. Using an arsenal of performance-analysis techniques including analytical models, custom microbenchmarks, full applications, and simulators, we succeeded in observing a serious-but previously undetected-performance problem. We identified the source of the problem, eliminated the problem, and "closed the loop" by demonstrating up to a factor of 2 improvement in application performance. We present our methodology and provide insight into performance analysis that is immediately applicable to other large-scale supercomputers.}, author = {Petrini, F. and Kerbyson, D. J. and Pakin, S.}, booktitle = {Supercomputing, 2003 ACM/IEEE Conference}, citeulike-article-id = {3393568}, citeulike-linkout-0 = {http://dx.doi.org/10.1109/SC.2003.10010}, citeulike-linkout-1 = {http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=1592958}, doi = {10.1109/SC.2003.10010}, journal = {Supercomputing, 2003 ACM/IEEE Conference}, keywords = {osnoise}, pages = {55}, posted-at = {2008-10-10 15:47:18}, priority = {2}, title = {The Case of the Missing Supercomputer Performance: Achieving Optimal Performance on the 8,192 Processors of ASCI Q}, url = {http://dx.doi.org/10.1109/SC.2003.10010}, year = {2003} }

TY - CONF ID - citeulike:3393568 L3 - citeulike-article-id:3393568 N2 - In this paper we describe how we improved the effective performance of ASCI Q, the world's second-fastest supercomputer, to meet our expectations. Using an arsenal of performance-analysis techniques including analytical models, custom microbenchmarks, full applications, and simulators, we succeeded in observing a serious-but previously undetected-performance problem. We identified the source of the problem, eliminated the problem, and "closed the loop" by demonstrating up to a factor of 2 improvement in application performance. We present our methodology and provide insight into performance analysis that is immediately applicable to other large-scale supercomputers. JF - Supercomputing, 2003 ACM/IEEE Conference EP - 55 TI - The Case of the Missing Supercomputer Performance: Achieving Optimal Performance on the 8,192 Processors of ASCI Q T2 - Supercomputing, 2003 ACM/IEEE Conference SP - 55 KW - osnoise AU - Petrini, F AU - Kerbyson, DJ AU - Pakin, S PY - 2003/// UR - http://dx.doi.org/10.1109/SC.2003.10010 ER -