Skip to content
Snippets Groups Projects

Repository graph

You can move around the graph by using the arrow keys.
Select Git revision
  • exclusive
  • main default protected
2 results
Created with Raphaël 2.2.02Feb131Oct3029262524Sep201918171615118Aug18Junmodified sh filemainmainmodified sh filereduced LBM single gpu codereduced LBM nvshmem codechange LBM mpi folder structureadd Copyrightchanged folder namemodified script filechange script file in nvshemem jacobichange chunk size compute methodchange file structure and nameclean redundant codeupdate dataMerged leonardo into alexupdate jacobi nvshmem longsome dataexclusiveexclusivelbm nvshmem 1.0LBN nvshmem add kerneladd nvtx to LBNadd nvtxadd new result of nvshmemadd new result of mpi-cudejacobi mpi strong and weakcleanup and weak scalingexclusive strong scaling over gpus and nodeschange configuration to run on Leonardodelete unnecessary fileEach gpu generate each part of outputinitial test passed, waiting for more testfinish LBM cuda aware mpi(2-8 GPUs)deal with int overflow problemreturn to ba9cec2ecombine all to timestepcombine propagate+rebound+collisionfix kernel indexcuda aware mpi finishmpi finish, final_state pass, av_vels doesn't passcuda aware mpi finish, but gpu increase, runtime doesn't reduceadd exchangeadd mpi local ny