並列分散プログラミング

I will cover the topics below, ranging from a very gentle introduction to parallel programming to performance of parallel programs There will be a few hands-on programming sessions that expose students to parallel programming and high performance programming experiences. * Parallel problem solving - It's easy! --- a quick and gentle introduction to parallel problem solving - Some examples of parallel problem solving and programming * Popular parallel programming models and their execution mechanisms - OpenMP, TBB, Cilk, etc. - parallel for and scheduling - tasks and dynamic load balancing - CUDA - warps and thread blocks - simultaneous multithreading - MPI * Parallelism : the source of performance - vectorization (SIMD) - superscalar parallelism - multicores and many cores - simultaneous multithreading (GPU) * Getting high performance - understanding the maximum performance of your CPU/GPU - understanding the maximum performance of memory - memory hierarchy - caches and high bandwidth memory

* lectures, * hands-on sessions (programming exercises) on real parallel machines

* attendance * reports

Students will work on high performance programming exercise on real parallel machines (multicores and GPUs). Basic experiences on C/C++ programming are recommended.