1. 30 Jan, 2018 1 commit
  2. 11 Jan, 2018 1 commit
  3. 13 Sep, 2017 2 commits
  4. 30 Aug, 2017 1 commit
  5. 19 Jul, 2017 1 commit
    • THIBAULT Samuel's avatar
      Reset optind to 1 before parsing options · 8bc688b6
      THIBAULT Samuel authored
      Normally the system is already doing this for us. In the MPI simgrid
      execution case, however, depending on the enabled optimizations for memory,
      libc may not be privatized, to save the memory that would be needed for its
      duplication, and to avoid the potential headaches brought by several libcs
      running in the same process. The first process would then inrease optind,
      and other processes would start with optind already increased.
      8bc688b6
  6. 06 Jul, 2017 13 commits
  7. 14 Apr, 2017 1 commit
    • THIBAULT Samuel's avatar
      Add Out-of-Core option · 3e6305c6
      THIBAULT Samuel authored
      Add MORSE_Desc_Create_OOC, which is like MORSE_Desc_Create, but does not
      actually allocate a matrix, thus letting the runtime allocate on-demand the
      tiles, possibly pushing them to the disk.
      
      Add a --ooc option to tests to enable this.
      3e6305c6
  8. 14 Mar, 2017 1 commit
  9. 09 Mar, 2017 1 commit
  10. 06 Mar, 2017 1 commit
  11. 14 Feb, 2017 1 commit
  12. 23 Dec, 2016 1 commit
  13. 09 Dec, 2016 1 commit
  14. 01 Dec, 2016 1 commit
    • PRUVOST Florent's avatar
      Re-work the cmake for simulation mode: · 7255c7c4
      PRUVOST Florent authored
          - use (starpu_cpu_func_t) 1 trick, same as cuda_func
          - cpu funtions are not defined anymore avoiding the dependency to coreblas
          - add #if !defined(CHAMELEON_SIMULATION) where it is needed
          - remove dependency to the coreblas library (become useless)
          - remove useless simucblas, simulapacke libraries
          - remove CHAMELEON_SIMULATION_MAGMA cmake variable and definition
            - keep using CHAMELEON_USE_CUDA and CHAMELEON_USE_MAGMA to consider CUDA kernels
            - this avoid to introduce useless new variables
          - work on messages
      
      7255c7c4
  15. 12 Oct, 2016 2 commits
  16. 20 Sep, 2016 2 commits
    • Guillaume Sylvand's avatar
      Add possibility to use z/cgemm3m for complex mat-mat products · 747c7935
      Guillaume Sylvand authored
      This routine, available in MKL, does a product in 6n^3 ops instead of 8n^3
      but is interesting only for "large enough" matrices (to be tested...)
      Potentially, we gain 25 % in all complex computations.
      It could be interesting to look for it / implement it in cuda.
      
      !!! Note that the flop counters are not updated         !!!
      !!! In C/Z accuracy, most flops counter should be x0.75 !!!
      
      IT is OFF by default
      It is activated with MORSE_Enable(MORSE_GEMM3M)
      In the timing routines, it is activated with --gemm3m
      747c7935
    • Guillaume Sylvand's avatar
      Add a 'progress indicator' feature, that displays a percentage of completion · 92a3c4a1
      Guillaume Sylvand authored
      IT is OFF by default
      It is activated with MORSE_Enable(MORSE_PROGRESS)
      In the timing routines, it is activated with --progress
      No progress is printed for tasks faster than 10 seconds
      92a3c4a1
  17. 09 Sep, 2016 1 commit
  18. 07 Sep, 2016 1 commit
  19. 05 Oct, 2015 1 commit
  20. 29 Sep, 2015 1 commit
  21. 28 Sep, 2015 1 commit
  22. 17 Sep, 2015 1 commit
  23. 16 Sep, 2015 1 commit
    • THIBAULT Samuel's avatar
      Introduce MORSE_Distributed_start, MORSE_Distributed_stop, · 34558c7a
      THIBAULT Samuel authored
      MORSE_Distributed_size, MORSE_Distributed_rank so that applications do not
      hardcode the use of MPI.
      
      Introduce RUNTIME_distributed_rank, RUNTIME_distributed_size,
      RUNTIME_distributed_barrier, so that MORSE does not hardcode the use of MPI
      either.
      
      This allows to use simgrid-mpi.
      34558c7a
  24. 28 Jul, 2015 1 commit
  25. 05 Feb, 2015 1 commit