Abstract
Performance and power consumption analysis and characterization for computational benchmarks is important for processor designers and benchmark developers. In this paper, we characterize and analyze different High Performance Computing workloads. We analyze benchmarks characteristics and behavior on various processors and propose a performance estimation analytical model to predict performance for different processor microarchitecture parameters. Performance model is verified to predict performance within <5% error margin between estimated and measured data for different processors. We also propose a power estimation analytical model to estimate power consumption with low error deviation.
Similar content being viewed by others
References
AMD K10- http://www.xbitlabs.com/articles/cpu/display/amd-k10.html
AMD Opteron-K8: http://www.cpu-world.com/CPUs/K8/index.html
Baghsorkhi S., Delahaye M., Patel S., Gropp W., Hwu W., An adaptive performance modeling tool for GPU architectures, In: Proceedings of ACM PPOPP, 105–114, 2010
Bhatai N., Alam S., Performance modeling of emerging HPC architectures, HPCMP Users Group Conference, June 2006
Chau-Yi Chou, A semi-empirical model for maximal LINPACK performance predictions”, 6th IEEE International Symposium on Cluster Computing and the Grid, 30 May 2006
Goel et al., Portable, scalable, per-core power estimation for intelligent resource management, Int. Green Computing Conference, 2010
Gustafon J., Todi R., Conventional Benchmarks as a sample of the Performance Spectrum, J. Super Comput., 13, 321–342, 1999
Hennessy J.L., Patterson D.A, Computer Architecture: A Quantitative Approach (4th Ed., Morgan Kaufmann, 2007)
http://www.nvidia.com/content/GTC/documents/SC09_Dongarra.pdf
Isci et al., Live, runtime phase monitoring and prediction on real systems with application to dynamic power management, Int. Symposium on Microarchitecture, 2006
Jens S., Performance Prediction on Benchmak Programs for Massively parallel Architectures, 10th Internaltion conference of High-Performance Computer (HPCS), June 1996
Kamil S., Power efficiency for high performance computing, IEEE International Symposium on Parallel and Distributed processing, June 2008
Livny M., Basney J., Raman R., Tannenbaum T., Mechanisms for High Throughput Computing, SPEEDUP J., 1997
Nvidia GTX460 http://www.nvidia.com/object/product-geforce-gtx-460-us.html
Nvidia GTX570 http://www.nvidia.com/object/product-geforce-gtx-570-us.html
Nvidia GTX580 http://www.nvidia.com/object/product-geforce-gtx-580-us.html
Nvidia nTune utility http://www.nvidia.com/object/ntune_2.00.23.html
Nvidia Tesla C2070 http://www.nvidia.com/object/personal-supercomputing.html
Rafael Saavedra H., Smith A.J., Analysis of benchmark characteristics and benchmark performance prediction, ACM Transactions on Comput. Syst., 14, 1996
Rohr D. et al., Multi-GPU DGEMM and High Performance LINPACK on Highly Energy-Efficient Clusters, IEEE Micro, September 2011
Ryoo S., Rodrigues C.I, Baghsorkhi S.S., Stone S.S., Kirk D.B., Hwu W.W., Optimization Principles and Application Performance Evaluation of a Multithreaded GPU using CUDA, In: Proceedings of the 13th ACM SIGPLAN, 73C82, ACM Press, 2008
Snavely A. et al., A Framework for Application Performance Modeling and Prediction, ACM/IEEE Supercomputing Conference, 2002
Volkov V., Demmel J.W., Benchmarking GPUs to Tune Dense Linear Algebra SC08, November 2008
Author information
Authors and Affiliations
Corresponding author
About this article
Cite this article
Issa, J. Performance and power analysis for high performance computation benchmarks. centr.eur.j.comp.sci. 3, 1–16 (2013). https://doi.org/10.2478/s13537-013-0101-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.2478/s13537-013-0101-5