Code

BSR block SpGEMM implementation#1099

Code

Merged

lucbv merged 16 commits intokokkos:developkokkos/kokkos-kernels:developfrom fnrizzi:29-bspgemm-implementationfnrizzi/kokkos-kernels:29-bspgemm-implementationCopy head branch name to clipboardApr 29, 2022

+4,650-52

Commits

Commits on Apr 22, 2022

• Copy SpGEMM files for reference diff

  

  Mikołaj Zuzek

  committed
  41070dc

  Copy full SHA for 41070dc

• Unit test for block SpGEMM

  

  Mikołaj Zuzek

  committed
  58959c7

  Copy full SHA for 58959c7

• ETI specializations for block SpGEMM

  

  Mikołaj Zuzek

  committed
  40e8d85

  Copy full SHA for 40e8d85

• Implementation of serial/debug block SpGEMM

  

  Mikołaj Zuzek

  committed
  736be46

  Copy full SHA for 736be46

• Implementation of default block SpGEMM algorithm

  

  Mikołaj Zuzek

  committed
  03ab278

  Copy full SHA for 03ab278

• fix kk_vector_block_add_mul() naming

  

  Mikołaj Zuzek

  committed
  dbb1f11

  Copy full SHA for dbb1f11

• clean up unused macros

  

  Mikołaj Zuzek

  committed
  346539f

  Copy full SHA for 346539f

• refactor view indexing

  

  Mikołaj Zuzek

  committed
  433f69c

  Copy full SHA for 433f69c

• fix literal max int

  

  Mikołaj Zuzek

  committed
  bd1e495

  Copy full SHA for bd1e495

• fix row_size naming

  

  Mikołaj Zuzek

  committed
  dd4b02c

  Copy full SHA for dd4b02c

• clean outdated comment

  

  Mikołaj Zuzek

  committed
  4c4e3c8

  Copy full SHA for 4c4e3c8

• refactor unused return value

  

  Mikołaj Zuzek

  committed
  3514cb3

  Copy full SHA for 3514cb3

• clean up unused code

  

  Mikołaj Zuzek

  committed
  682a175

  Copy full SHA for 682a175

• Clean SpGEMM code not used in block version

  

  Mikołaj Zuzek

  committed
  e380d1b

  Copy full SHA for e380d1b

• Disable unavailable TPL implementations

  

  Mikołaj Zuzek

  committed
  dcc1600

  Copy full SHA for dcc1600

• add some explanation to CPU functor in speed method

  

  Mikołaj Zuzek

  committed
  b712a56

  Copy full SHA for b712a56