Model-Based Relative Performance Diagnosis of Wavefront Parallel Computations

作者: Li Li , Allen D. Malony , Kevin Huck

DOI: 10.1007/11847366_21

关键词:

摘要: Parallel performance diagnosis can be improved with the use of knowledge about parallel computation models. The Hercule system applies model-based methods to automate processes and explain problems from high-level semantics. However, is limited by a single experiment view. Here we introduce concept relative show how it integrated in framework. paper demonstrates effectiveness Hercule's approach well-known Sweep3D application based on Wavefront model. Relative diagnoses anomalies strong weak scaling cases are given.

参考文章(11)
Rok Sosic, Ian T. Foster, David Abramson, John G. Michalakes, Relative Debugging: A new paradigm for debugging scientific applications Communications of The ACM. ,(1996)
Karen L. Karavanic, Barton P. Miller, A framework for multi-execution performance tuning On-line monitoring systems and computer tool interoperability. pp. 61- 89 ,(2004)
David Abramson, Ian Foster, John Michalakes, Rok Sosič, Relative debugging Communications of the ACM. ,vol. 39, pp. 69- 77 ,(1996) , 10.1145/240455.240475
B.P. Miller, M.D. Callaghan, J.M. Cargille, J.K. Hollingsworth, R.B. Irvin, K.L. Karavanic, K. Kunchithapadam, T. Newhall, The Paradyn parallel performance measurement tool IEEE Computer. ,vol. 28, pp. 37- 46 ,(1995) , 10.1109/2.471178
J. Dongarra, F. Song, S. Moore, F. Wolf, N. Bhatia, An algebra for cross-experiment performance analysis international conference on parallel processing. pp. 63- 72 ,(2004) , 10.1109/ICPP.2004.15
K.L. Karavanic, J. May, K. Mohror, B. Miller, K. Huck, R. Knapp, B. Pugh, Integrating Database Technology with Comparison-based Parallel Performance Diagnosis: The PerfTrack Performance Experiment Management Tool conference on high performance computing (supercomputing). pp. 39- 39 ,(2005) , 10.1109/SC.2005.36
K.A. Huck, A.D. Malony, PerfExplorer: A Performance Data Mining Framework For Large-Scale Parallel Computing conference on high performance computing (supercomputing). pp. 41- 41 ,(2005) , 10.1109/SC.2005.55
L. Li, A. D. Malony, Knowledge engineering for automatic parallel performance diagnosis Concurrency and Computation: Practice and Experience. ,vol. 19, pp. 1497- 1515 ,(2007) , 10.1002/CPE.1127
Karen L. Karavanic, Barton P. Miller, Improving Online Performance Diagnosis by the Use of Historical Performance Data conference on high performance computing (supercomputing). pp. 42- 42 ,(1999) , 10.1145/331532.331574
Philip C. Roth, Barton P. Miller, On-line automated performance diagnosis on thousands of processes acm sigplan symposium on principles and practice of parallel programming. pp. 69- 80 ,(2006) , 10.1145/1122971.1122984