Scientific Computing with Multicore and Accelerators

摘要: The hybrid/heterogeneous nature of future microprocessors and large high-performance computing systems will result in a reliance on two major types components: multicore/manycore central processing units special purpose hardware/massively parallel accelerators. While these technologies have numerous benefits, they also pose substantial performance challenges for developers, including scalability, software tuning, programming issues. Researchers at the Forefront Reveal Results from Their Own State-of-the-Art WorkEdited by some top researchers field with contributions variety international experts, Scientific Computing Multicore Accelerators focuses architectural design implementation multicore manycore processors accelerators, graphics (GPUs) Sony Toshiba IBM (STI) Cell Broadband Engine (BE) currently used PlayStation 3. book explains how numerical libraries, such as LAPACK, help solve computational science problems; explores emerging area hardware-oriented numerics; presents fast Fourier transform (FFT) list ranking algorithm BE. It covers stencil computations, auto-tuning, optimizations kernel, sequence alignment homology, pairwise computations. evaluates portability drug applications to BE illustrates successfully exploit capabilities GPUs scientific applications. concludes chapters dataflow frameworks, Charm++ model, scan algorithms, portable intracore communication framework. Explores New Computational Landscape Hybrid Processors By offering insight into process constructing effectively using technology, this volume provides thorough practical introduction hybrid computing. discusses introductory concepts simple examples computing, logical debugging advanced topics issues related use building many