HPGMG implements full multigrid (FMG) algorithms using finite-volume and finite-element methods. Different algorithmic variants adjust the arithmetic intensity and architectural properties that are tested. It is a benchmark for HPC machines based on geometric multigrid methods. Both are full multigrid (FMG) methods using Chebyshev smoothers, thus have similar large-scale communication patterns, but the local computational kernels and memory/cache demands are different for each discretization.


HPGMG has been ported with  the  HCC Compiler with the C++ AMP dialect which support

  • C++ AMP version can be built to run on
    • FirePro W9100 and S9150 via OpenCL 1.2 runtime with SPIR 1.2
    • Radeon Open Compute Kernel Driver and Runtime on Fiji based hardware.

You can get HPGMG GitHub linked bellow.