A restructuring method for implementing numerical algorithms for scientific computing can help them run efficiently on the IBM Cell processor and other multicore CPUs. Using the PPM gas dynamics algorithm as an example, this work walks through code transformation techniques step by step that can boost performance of scientific simulation codes.
Index Terms:
high performance computing, multicore computation, Cell processor, program transformations for performance
Citation:
Paul R. Woodward, Jagan Jayaraj, Pei-Hung Lin, Pen-Chung Yew, "Moving Scientific Codes to Multicore Microprocessor CPUs," Computing in Science and Engineering, vol. 10, no. 6, pp. 16-25, Nov./Dec. 2008, doi:10.1109/MCSE.2008.152