Automatic mapping of Fortran programs onto clusters with graphics processing units
Автор: Bakhtin V.A., Klinov M.S., Kolganov A.S., Krukov V.A., Podderyugina N.V., Pritula M.N.
Статья в выпуске: 3 т.3, 2014 года.
Бесплатный доступ
The paper shows results of exploitation of automated parallelization system SAPFOR for sequential programs parallelization for cluster with GPUs, including programs with regular data de-pendencies. The system translates Fortran program into Fortran DVMH program. Produced pro-gram can be executed on cluster. Fortran DVMH language, its compilers and debugging tools are bundled within DVM-system. Performed transformations with original programs are considered. Programs which use a variety of parallel programming technologies are emitted by the system. Source code characteristics are considered. Performance analysis is done on CPUs and GPUs of the K-100 computational cluster.
Dvm for heterogeneous systems, fortran dvmh, fortran, openmp, hybrid computational systems with accelerators, gpu
Короткий адрес: https://sciup.org/147160542
IDR: 147160542