1. 09 Jul, 2021 2 commits
  2. 10 Jun, 2021 1 commit
  3. 20 Apr, 2021 1 commit
  4. 15 Apr, 2021 1 commit
  5. 12 Mar, 2021 2 commits
    • Andreas Marek's avatar
      Forgot to set rocBlas handle · 979c5985
      Andreas Marek authored
      With this commit ELPA does run correctly on AMD GPUs.
      Functionality and correctness tests have been carried out
      on AMD MI100
      979c5985
    • Andreas Marek's avatar
      Forgot to set rocBlas handle · 4a4d6051
      Andreas Marek authored
      With this commit ELPA does run correctly on AMD GPUs.
      Functionality and correctness tests have been carried out
      on AMD MI100
      4a4d6051
  6. 04 Mar, 2021 1 commit
  7. 03 Mar, 2021 1 commit
  8. 02 Mar, 2021 1 commit
  9. 27 Feb, 2021 2 commits
  10. 26 Feb, 2021 4 commits
  11. 25 Feb, 2021 1 commit
  12. 24 Feb, 2021 3 commits
  13. 15 Feb, 2021 1 commit
  14. 14 Dec, 2020 1 commit
  15. 18 Nov, 2020 1 commit
  16. 17 Nov, 2020 1 commit
  17. 25 Sep, 2020 1 commit
  18. 18 Sep, 2020 1 commit
  19. 05 Jun, 2020 1 commit
  20. 02 Jun, 2020 1 commit
  21. 20 Nov, 2019 2 commits
    • Wenzhe Yu's avatar
      Extend CUDA wrapper · 6e5c03a6
      Wenzhe Yu authored
      * cudaMallocHost
      * cudaFreeHost
      * cudaHostRegister
      * cudaHostUnregister
      6e5c03a6
    • Wenzhe Yu's avatar
      Rewrite compute_hh_trafo CUDA kernels · 6cd5a4f1
      Wenzhe Yu authored
      * Switch to a simple non-WY algorithm
      * Unify real and complex cases
      * Update reduction kernel
      * Use __shfl_xor_sync for warp reduce (CUDA 9+)
      * Support 2^n block size, n = 1,2,...,10
      * Use templates when possible
      * Clean up unused CUDA functions
      * Increase default stripe width when using GPU
      6cd5a4f1
  22. 30 Oct, 2019 1 commit
  23. 23 Oct, 2019 1 commit
  24. 22 Oct, 2019 2 commits
    • Sebastian Ohlmann's avatar
      Add support for NVTX profiling · bf7c6410
      Sebastian Ohlmann authored
      When profiling the GPU version, NVTX can be used to highlight the
      corresponding regions of the code in the timeline of the profiling tool
      (nvvp or nsight systems). This is very useful to correlate what happens
      on the GPU with what part of the code we are in.
      bf7c6410
    • Pavel Kus's avatar
      adding tool to check GPU memory · 82ffbd38
      Pavel Kus authored
      82ffbd38
  25. 14 Oct, 2019 1 commit
  26. 15 Feb, 2019 1 commit
  27. 15 Oct, 2018 1 commit
    • Pavel Kus's avatar
      doing GPU initialization for the first time only · d900a3e1
      Pavel Kus authored
      The GPU initialization is actually quite constly, e.g. on Minsky it
      takes roughly 0.7s. That is hurting performance for small matrices.
      Thus a check has been added and now GPU should be initialized only the
      first time.
      d900a3e1
  28. 22 Feb, 2018 1 commit
  29. 11 Sep, 2017 1 commit
  30. 29 Aug, 2017 1 commit