View Document Preview and Link Document Date: 2008-07-27 14:02:45 Open Document File Size: 203,12 KB Share Result on Facebook
City P. Hanrahan / Houston / / Company Los Alamos National Laboratory / PeakStream / NVIDIA Corporation / Advanced Computing Systems / HIOKI E.E. CORPORATION / SPRAT / J. Kruger A. E. / Intel / / / Facility Tokyo Institute of Technology / Morgan Kaufmann Pub / Tohoku University / / IndustryTerm purpose applications / heterogeneous computing resources / higher-level programming tools / stream processing task / co-processors / stream processing manner / parallel processing applications / appropriate computing engine / cell processor / appropriate processor / dense linear systems / numerical and scientific applications / stream processing application / software development / energy / hybrid computing systems / accumulated energy difference / heterogeneous multicore processors / hybrid computing system / energy consumption / widelyavailable hybrid computing system / stream processing applications / graphics processors / stream processing tasks / graphics hardware / generalpurpose processors / gpgpu applications / virtual computing system / energy saving / scientific algorithms / energy efficiency / computing / energy-aware computing / energyaware computing / stream processing / visualization tool / data-parallel co-processor / energydelay product / / NaturalFeature argv){ stream / char** argv){ stream / / OperatingSystem Linux / CentOS / / Organization university of Tokyo / Tohoku University / Graduate School / Cenberscience Center / Tokyo Institute of Technology / / Person Hiroyuki Takizawa / Hiroaki Kobayashi / Tq / Reiji Suda / Kentaro Sano / / / Position driver / map qualifier / performance model for assisting development / memory model for scientific algorithms / kernel qualifier / programmer / / Product GF88GTX / HiTESTER 3334 / TC / / ProgrammingLanguage C / C++ / / ProvinceOrState British Columbia / South Carolina / / Technology generalpurpose processors / Broadband / SMs / graphics processors / Linux / API / Runtime Processor / cache memory / fluid dynamics / random access / html / cell processor / data-parallel co-processor / one processor / appropriate processor / key technology / http / heterogeneous multicore processors / simulation / parallel processing / / URL http / SocialTag 
SPRAT: Runtime Processor Selection for Energy-aware Computing Hiroyuki Takizawa #1 , Katuto Sato #2 , and Hiroaki Kobayashi ∗3 # Graduate School of Information Sciences, Tohoku University