Evaluating multi-core platforms for HPC data-intensive kernels

ABSTRACT

Multi-core platforms have proven themselves able to accelerate numerous HPC applications. But programming data-intensive applications on such platforms is a hard, and not yet solved, problem. Not only do modern processors favor compute-intensive code, they also have diverse architectures and incompatible programming models. And even after making a difficult platform choice, extensive programming effort must be invested with an uncertain performance outcome. By taking the plunge on an irregular, data-intensive application, we present an evaluation of three platform types, namely the generic multi-core CPU, the STI Cell/B.E., and the GPU. We evaluate these platforms in terms of application performance, programming effort and cost. Although we do not select a clear winner, we do provide a list of guidelines to assist in platform choice and development of similar data-intensive applications.

REFERENCES

• 1: David A. Bader , Virat Agarwal , Kamesh Madduri , Seunghwa Kang, High performance combinatorial algorithm design on the Cell Broadband Engine processor, Parallel Computing, v.33 n.10-11, p.720-740, November, 2007 [doi>10.1016/j.parco.2007.09.005]
• 2: C. Benthin, I. Wald, M. Scherbaum, and H. Friedrich. Ray tracing on the Cell processor. In IEEE Symposium of Interactive Ray Tracing, pages 15--23. IEEE Computer Society Press, September 2006.
• 3: F. Blagojevic, A. Stamatakis, C. Antonopoulos, and D.S. Nikolopoulos. RAxML-CELL: Parallel phylogenetic tree construction on the Cell Broadband Engine. In IEEE International Parallel and Distributed Processing Symposium. IEEE Press, March 2007.
• 4: Arndt Bode, High Performance Computing in the Multi-core Area, Proceedings of the Sixth International Symposium on Parallel and Distributed Computing, p.4, July 05-08, 2007 [doi>10.1109/ISPDC.2007.28]
• 5: D.S. Briggs, F.R. Schwab, and R.A. Sramek. Imaging. In Synthesis Imaging in Radio Astronomy II, volume 180 of Astronomical Society of the Pacific Conference Series, pages 127--140, 1999.
• 6: A. Buttari, P. Luszczek, J. Kurzak, J. Dongarra, and G. Bosilca. SCOP3: A rough guide to scientific computing on the PlayStation 3. Technical Report UT-CS-07-595, Innovative Computing Lab., University of Tennessee, Knoxville, April 2007.
• 7: Theofanis Constantinou , Yiannakis Sazeides , Pierre Michaud , Damien Fetis , Andre Seznec, Performance implications of single thread migration on a chip multi-core, ACM SIGARCH Computer Architecture News, v.33 n.4, November 2005 [doi>10.1145/1105734.1105745]
• 8: T. Cornwell, K. Golap, and S. Bhatnagar. W projection: A new algorithm for wide field imaging with radio synthesis arrays. In Astronomical Data Analysis Software and Systems XIV, volume 347, pages 86--95. ASP Press, October 2004.
• 9: T.J. Cornwell. SKA and EVLA computing costs for wide field imaging. Experimental Astronomy, 17:329--343, June 2004.
• 10: T.J. Cornwell and R.A. Perley. Radio-interferometric imaging of very large fields -- The problem of non-coplanar arrays. Astronomy and Astrophysics, 261:353--364, July 1992.
• 11: Kaushik Datta , Mark Murphy , Vasily Volkov , Samuel Williams , Jonathan Carter , Leonid Oliker , David Patterson , John Shalf , Katherine Yelick, Stencil computation optimization and auto-tuning on state-of-the-art multicore architectures, Proceedings of the 2008 ACM/IEEE conference on Supercomputing, November 15-21, 2008, Austin, Texas
• 12: U. Drepper. What Every Programmer Should Know About Memory, November 2007.
• 13: Kayvon Fatahalian , Daniel Reiter Horn , Timothy J. Knight , Larkhoon Leem , Mike Houston , Ji Young Park , Mattan Erez , Manman Ren , Alex Aiken , William J. Dally , Pat Hanrahan, Sequoia: programming the memory hierarchy, Proceedings of the 2006 ACM/IEEE conference on Supercomputing, November 11-17, 2006, Tampa, Florida [doi>10.1145/1188455.1188543]
• 14: Khronos OpenCL Working Group. OpenCL 1.0 Standard, December 2008.
• 15: Michael Kistler , Michael Perrone , Fabrizio Petrini, Cell Multiprocessor Communication Network: Built for Speed, IEEE Micro, v.26 n.3, p.10-23, May 2006 [doi>10.1109/MM.2006.49]
• 16: Jens Krüger , Rüdiger Westermann, Linear algebra operators for GPU implementation of numerical algorithms, ACM Transactions on Graphics (TOG), v.22 n.3, July 2003
• 17: L.-K. Liu, Q. Liu, A.P. Natsev, K.A. Ross, J.R. Smith, and A.L. Varbanescu. Digital media indexing on the Cell processor. In IEEE International Conference on Multimedia and Expo, pages 1866--1869, July 2007.
• 18: W. Liu, B. Schmidt, G. Voss, and W. Müller-Wittig. Molecular dynamics simulations on commodity GPUs with CUDA. In High Performance Computing, pages 185--196, December 2007.
• 19: M. McCool. Signal processing and general-purpose computing on GPUs. IEEE Signal Processing Magazine, pages 109--114, May 2007.
• 20: S.K. Moore. Multicore is bad news for supercomputers. IEEE Spectrum, November 2008.
• 21: NVIDIA. CUDA Programming Guide, December 2008.
• 22: F. Petrini, J. Fernàndez, M. Kistler, G. Fossum, A.L. Varbanescu, and M. Perrone. Multicore surprises: Lessons learned from optimizing Sweep3D on the Cell Broadband Engine. In IEEE International Parallel and Distributed Processing Symposium. IEEE Press, March 2007.
• 23: S. S. Stone , J. P. Haldar , S. C. Tsao , W. -m. W. Hwu , B. P. Sutton , Z. -P. Liang, Accelerating advanced MRI reconstructions on GPUs, Journal of Parallel and Distributed Computing, v.68 n.10, p.1307-1318, October, 2008 [doi>10.1016/j.jpdc.2008.05.013]
• 24: K. van der Schaaf, C. Broekema, G. Diepen, and E. Meijeren. The LOFAR central processing facility architecture. Experimental Astronomy, 17(1-3):43--58, June 2004.
• 25: R.V. van Nieuwpoort and J.W. Romein. Using many-core hardware to correlate radio astronomy signals. In 23rd ACM International Conference on Supercomputing, June 2009, to appear.
• 26: A.L. Varbanescu, A. van Amesfoort, T. Cornwell, B.G. Elmegreen, R. van Nieuwpoort, G. van Diepen, and H. Sips. The performance of gridding/degridding on the Cell/B.E. Technical report, Delft University of Technology, January 2008.
• 27: A.L. Varbanescu, A. van Amesfoort, T. Cornwell, B.G. Elmegreen, R. van Nieuwpoort, G. van Diepen, and H. Sips. Radioastronomy image synthesis on the Cell/B.E. Technical report, Delft University of Technology, August 2008.
• 28: R. Wayth, K. Dale, L. Greenhill, D. Mitchell, S. Ord, and H. Pfister. Real-time calibration and imaging for the MWA (poster). In AstroGPU, November 2007.
• 29: S. Williams, J. Carter, L. Oliker, J. Shalf, and K.A. Yelick. Lattice boltzmann simulation optimization on leading multicore platforms. In Symposium on Parallel and Distributed Processing, pages 1--14, April 2008.
• 30: Samuel Williams , John Shalf , Leonid Oliker , Shoaib Kamil , Parry Husbands , Katherine Yelick, The potential of the cell processor for scientific computing, Proceedings of the 3rd conference on Computing frontiers, May 03-05, 2006, Ischia, Italy [doi>10.1145/1128022.1128027]
• 31: Samuel Williams , Andrew Waterman , David Patterson, Roofline: an insightful visual performance model for multicore architectures, Communications of the ACM, v.52 n.4, April 2009 [doi>10.1145/1498765.1498785]

INDEX TERMS

Primary Classification:
  C. Computer Systems Organization
  C.1 PROCESSOR ARCHITECTURES
      C.1.3 Other Architecture Styles
          Subjects: Heterogeneous (hybrid) systems

Additional Classification:
  C. Computer Systems Organization
  C.1 PROCESSOR ARCHITECTURES
      C.1.4 Parallel Architectures
  C.4 PERFORMANCE OF SYSTEMS
      Subjects: Performance attributes

  D. Software
  D.1 PROGRAMMING TECHNIQUES
      D.1.3 Concurrent Programming
          Subjects: Parallel programming

General Terms:
Algorithms, Measurement, Performance

Keywords:
data-intensive kernels, gpus, memory-bound applications, multi-core processors, the cell processor