Abstract
The High Performance Conjugate Gradient (HPCG) benchmark has been recently proposed as a complement to the High Performance Linpack (HPL) benchmark currently used to rank supercomputers in the Top500 list. This new benchmark solves a large sparse linear system using a multigrid preconditioned conjugate gradient (PCG) algorithm. The PCG algorithm contains the computational and communication patterns prevalent in the numerical solution of partial differential equations and is designed to better represent modern application workloads which rely more heavily on memory system and network performance than HPL. GPU accelerated supercomputers have proved to be very effective, especially with regard to power efficiency, for accelerating compute intensive applications like HPL. This paper will present the details of a CUDA implementation of HPCG, and the results obtained at full scale on the largest GPU supercomputers available: the Cray XK7 at ORNL and the Cray XC30 at CSCS. The results indicate that GPU accelerated supercomputers are also very effective for this type of workload.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Dongarra, J., Heroux, M.A.: Toward a New Metric for Ranking High Performance Computing Systems. Sandia report SAND2013-4744 (2013)
Dongarra, J., Luszczek, P.: Introduction to the HPC challenge benchmark Suite, ICL Technical report, ICL-UT-05-01, (Also appears as CS Dept. Tech report UT-CS-05-544) (2005)
Heroux, M.A., Dongarra, J., Luszczek, P: HPCG Technical specification, Sandia report SAND2013-8752 (2013)
Graph 500. http://www.graph500.org
Green 500. http://www.green500.org
CUDA Toolkit. http://developer.nvidia.com/cuda-toolkit
CUDA Fortran. http://www.pgroup.com/resources/cudafortran.htm
CUBLAS Library. http://docs.nvidia.com/cuda/cublas
CUSPARSE Library. http://docs.nvidia.com/cuda/cusparse
THRUST Library. http://docs.nvidia.com/cuda/thrust
Barrett, R.F., Heroux, M.A., Lin, P.T., Vaughan, C.T., Williams, A.B.: Poster: mini-applications: vehicles for co-design. In: Proceedings of the 2011 Companion on High Performance Computing Networking, Storage and Analysis Companion (SC 2011 Companion), pp. 1–2. ACM, New York (2011)
Golub, G.H., Van Loan, C.F.: Matrix Computations, 3rd edn. John Hopkins University Press, USA (1996)
Briggs, W.L., Henson, V.E., McCormick, S.F.: A Multigrid Tutorial. SIAM, USA (2000)
Green 500: Energy efficient HPC System workloads power measurement methodology (2013)
McCalpin, J.D.: Memory bandwidth and machine balance in current high performance computers. In: IEEE Computer Society Technical Committee on Computer Architecture (TCCA) Newsletter, December 1995
Phillips, E.H., Fatica, M.: Implementing the Himeno benchmark with CUDA on GPU clusters. In: IEEE International Symposium on Parallel & Distributed Processing IPDPS, pp. 1–10 (2010)
Park, J., Smelyanskiy, M.: Optimizing Gauss-Seidel Smoother in HPCG. In: ASCR HPCG Workshop, Bethesda MD, 25 March 2014
Luby, M.: A simple parallel algorithm for the maximal independent set problem. SIAM J. Comput. 15(4), 1036–1053 (1986)
Jones, M.T., Plassmann, P.E.: A parallel graph coloring heuristic. SIAM J. Sci. Comput. 14, 654–669 (1992)
Cohen, J., Castonguay, P.: Efficient graph matching and coloring on the GPU. In: GPU Technology Conference, San Jose CA, 14–17 May 2012. http://ondemand.gputechconf.com/gtc/2012/presentations/S0332-Efficient-Graph-Matching-and-Coloring-on-GPUs.pdf
Acknowledgments
This research used resources of the Oak Ridge Leadership Computing Facility at the Oak Ridge National Laboratory, which is supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC05-00OR22725. We wish to thank Buddy Bland, Jack Wells and Don Maxwell of Oak Ridge National Laboratory for their support. This work was also supported by a grant from the Swiss National Supercomputing Centre (CSCS) under project ID g33. We also want to acknowledged the support from Gilles Fourestey and Thomas Schulthess at CSCS. We wish to thank Lung Scheng Chien and Jonathan Cohen at NVIDIA for relevant discussions.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2015 Springer International Publishing Switzerland
About this paper
Cite this paper
Phillips, E., Fatica, M. (2015). A CUDA Implementation of the High Performance Conjugate Gradient Benchmark. In: Jarvis, S., Wright, S., Hammond, S. (eds) High Performance Computing Systems. Performance Modeling, Benchmarking, and Simulation. PMBS 2014. Lecture Notes in Computer Science(), vol 8966. Springer, Cham. https://doi.org/10.1007/978-3-319-17248-4_4
Download citation
DOI: https://doi.org/10.1007/978-3-319-17248-4_4
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-17247-7
Online ISBN: 978-3-319-17248-4
eBook Packages: Computer ScienceComputer Science (R0)