Performance analysis and visualization tools for cell/B.E. multicore environment

ABSTRACT

Code porting, optimizing and tuning has become a challenging task in multicore/many cores environment. It requires a different set of performance visualization tools to handle the complexity of the many cores and the size of performance data to find opportunities for optimization. This paper discusses performance visualization tools available for Cell/B.E. under the IBM Software Development Kit (SDK) for Multicore Acceleration Version 3.0. It also presents a methodology for porting, optimizing and tuning Cell applications by utilizing those tools. The paper starts with a simple scalar program example which can also be found in the IBM tutorial for the Cell programming, and then describes all the needed steps to make it fully tuned and scaled for the Cell Broadband Engine.

REFERENCES

• 1: Introductory overview of the Cell multiprocessor, <u>IBM JRD 49-4/5 | Introduction to the Cell multiprocessor</u>
• 2: A. E. Eichenberger , J. K. O'Brien , K. M. O'Brien , P. Wu , T. Chen , P. H. Oden , D. A. Prener , J. C. Shepherd , B. So , Z. Sura , A. Wang , T. Zhang , P. Zhao , M. K. Gschwind , R. Archambault , Y. Gao , R. Koo, Using advanced compiler technology to exploit the performance of the Cell Broadband Engine^TM architecture, IBM Systems Journal, v.45 n.1, p.59-84, January 2006
• 3: Introduction to the Cell Broadband Engine, <u>Cell Broadband Engine - IBM Microelectronics</u>
• 4: <u>Cell Broadband Engine#8482; processor -- based systems White Paper</u>
• 5: <u>Spufs: The Cell Synergistic Processing Unit as a virtual file system</u>
• 6: <u>"Maximizing the power of the Cell Broadband Engine processor: 25 tips to optimal application performance"</u>, by Dan Brokenshire
• 7: <u>FFTC: Fastest Fourier Transform for the IBM Cell Broadband Engine</u>, David A. Bader and Virat Agarwal College of Computing Georgia Institute of Technology, Atlanta, GA, USA 30332
• 8: Samuel Williams et al., <u>The Potential of the Cell Processor for Scientific Computing</u>, Computational Research Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94720
• 9: David A. Bader and Virat Agarwal, <u>FFTC: Fastest Fourier Transform for the IBM Cell Broadband Engine</u>, College of Computing Georgia Institute of Technology, Atlanta, GA, USA 30332
• 10: A. C. Chow, G. C. Fossum, D. A. Brokenshire, A Programming Example: Large Fast Fourier Transform on the Cell Broadband Engine. IBM White Paper at http://www-01.ibm.com/chips/techlib/techlib.nsf/techdocs/0AA2394A505EF0FB872570AB005BF0F1/$file/GSPx_FFT_paper_legal_0115.pdf
• 11: T. Chen, H. Mathis, <u>IBM BladeCenter QS20 Performance Potentials</u>, White paper by IBM corp. at: ftp://ftp.software.ibm.com/common/ssi/sa/wh/n/blw03001usen/BLW03001USEN.PDF
• 12: Dan Brokenshire, <u>"Maximizing the power of the Cell Broadband Engine processor: 25 tips to optimal application performance"</u>
• 13: Cell BE programming Tutorial http://www-01.ibm.com/chips/techlib/techlib.nsf/techdocs/FC857AE550F7EB83872571A80061F788?Open&S_TACT=105AGX16 &S_CMP=LP
• 14: Thomas Chen, et al., "Cell Broadband Engine Architecture and its First Implementation -- a Performance View", http://www.ibm.com/developerworks/power/library/pacellperf/
• 15: Samuel Williams, et al., "The Potential of the Cell Processor for Scientific Computing", http://www.cs.berkeley.edu/~samw/projects/cell/CF06.pdf#search=%22Cell%20performance%20%20FFT%22
• 16: Carsten Benthin, et al., "Ray Tracing on the Cell Processor", http://graphics.cs.unisb.de/~benthin/cellrt06.pdf#search=%22InTrace%20Ray%20Tracing%20Performance%20Opteron%22 bltiQS20perfwp103106.doc Page 5
• 17: Alex Chunghen Chow, et al., "A Programming Example: Large FFT on the Cell Broadband Engine", http://www.ibm.com/chips/techlib/techlib.nsf/techdocs/0AA2394A505EF0FB872570AB005BF0F1/$file/GSPx_FFT_paper_legal_0115.pdf#search=%22Cell%20performance%20%2 0FFT%22
• 18: B. Minor, G. Fossum, V. To, "Terrain Rendering Engine (TRE): Cell Broadband Optimized Real-Time Ray-Caster", Proceedings of GPSx conference, October 2005.
• 19: Sun Studio, http://developers.sun.com/sunstudio/
• 20: IntelVTune#8482; Performance Analyzer, http://www.intel.com/cd/software/products/asmona/eng/239144.htm
• 21: Cray PAT, http://nccs.gov/computing-resources/jaguar/debugging-optimization/cray-pat/
• 22: KOJAK, http://icl.cs.utk.edu/kojak/
• 23: SCALASCA, http://www.fz-juelich.de/jsc/scalasca/
• 24: Vampir, http://www.vampir.eu/
• 25: TAU, http://www.cs.uoregon.edu/research/tau/home.php
• 26: OProfile, http://oprofile.sourceforge.net
• 27: PDT, http://www.ibm.com/developerworks/edu/pa-dw-pasdk3tool.html
• 28: CPC, http://www.ibm.com/developerworks/edu/pa-dw-pasdk3tool.html
• 29: FDPR-Pro, http://www.haifa.ibm.com/projects/systems/cot/fdpr/
• 30: System Simulator, http://www.alphaworks.ibm.com/tech/cellsystemsim
• 31: VPA, http://www.alphaworks.ibm.com/tech/vpa
• 32: AMD Code Analyst, http://www.amd.com/usen/Processors/DevelopWithAMD/0,30_2252_869_3604%5 E9713,00.html
• 33: Paraver, http://www.cepba.upc.es/paraver/
• 34: Euler example for Cell, http://sysrun.haifa.il.ibm.com/hrl/euler_example_for_cell/
• 35: G. Fossum, Cell code porting, Internal IBM seminar.

INDEX TERMS

Primary Classification:
  C. Computer Systems Organization
  C.1 PROCESSOR ARCHITECTURES
      C.1.2 Multiple Data Stream Architectures (Multiprocessors)
          Subjects: Single-instruction-stream, multiple-data-stream processors (SIMD)

Additional Classification:
  C. Computer Systems Organization
  C.4 PERFORMANCE OF SYSTEMS
      Subjects: Measurement techniques

  D. Software
  D.4 OPERATING SYSTEMS
      D.4.8 Performance
          Subjects: Measurements

General Terms:
Algorithms, Measurement, Performance

Keywords:
Euler based particle simulation, cell broadband engine, code analyzer, control flow analyzer, counter analyzer, pipeline analyzer, profile analyzer, trace analyzer, virtual performance analyzer