From 8ea24215c156c1e1957960b31ba720147f96e66d Mon Sep 17 00:00:00 2001 From: Evgeny Mankov Date: Wed, 6 Nov 2024 21:38:05 +0100 Subject: [PATCH] [HIPIFY][BLAS] Sync with `CUDA 12.6.2` - Step 13 - BlasLt API + Updated the regenerated `hipify-perl` and `BLAS` `CUDA2HIP` docs accordingly --- bin/hipify-perl | 84 ++++++++ docs/tables/CUBLAS_API_supported_by_HIP.md | 42 ++++ .../CUBLAS_API_supported_by_HIP_and_ROC.md | 42 ++++ docs/tables/CUBLAS_API_supported_by_ROC.md | 42 ++++ src/CUDA2HIP_BLAS_API_types.cpp | 180 +++++++++++++----- 5 files changed, 342 insertions(+), 48 deletions(-) diff --git a/bin/hipify-perl b/bin/hipify-perl index 64f19bb9..8a4afeb6 100755 --- a/bin/hipify-perl +++ b/bin/hipify-perl @@ -12800,6 +12800,7 @@ sub warnHipOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_408x64", "CUBLASLT_MATMUL_TILE_400x64", "CUBLASLT_MATMUL_TILE_392x64", + "CUBLASLT_MATMUL_TILE_384x8", "CUBLASLT_MATMUL_TILE_384x64", "CUBLASLT_MATMUL_TILE_384x128", "CUBLASLT_MATMUL_TILE_376x64", @@ -12831,8 +12832,31 @@ sub warnHipOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_32x128", "CUBLASLT_MATMUL_TILE_328x64", "CUBLASLT_MATMUL_TILE_328x128", + "CUBLASLT_MATMUL_TILE_320x96", + "CUBLASLT_MATMUL_TILE_320x88", + "CUBLASLT_MATMUL_TILE_320x80", + "CUBLASLT_MATMUL_TILE_320x8", + "CUBLASLT_MATMUL_TILE_320x72", "CUBLASLT_MATMUL_TILE_320x64", + "CUBLASLT_MATMUL_TILE_320x56", + "CUBLASLT_MATMUL_TILE_320x48", + "CUBLASLT_MATMUL_TILE_320x40", + "CUBLASLT_MATMUL_TILE_320x32", + "CUBLASLT_MATMUL_TILE_320x24", + "CUBLASLT_MATMUL_TILE_320x200", + "CUBLASLT_MATMUL_TILE_320x192", + "CUBLASLT_MATMUL_TILE_320x184", + "CUBLASLT_MATMUL_TILE_320x176", + "CUBLASLT_MATMUL_TILE_320x168", + "CUBLASLT_MATMUL_TILE_320x160", + "CUBLASLT_MATMUL_TILE_320x16", + "CUBLASLT_MATMUL_TILE_320x152", + "CUBLASLT_MATMUL_TILE_320x144", + "CUBLASLT_MATMUL_TILE_320x136", "CUBLASLT_MATMUL_TILE_320x128", + "CUBLASLT_MATMUL_TILE_320x120", + "CUBLASLT_MATMUL_TILE_320x112", + "CUBLASLT_MATMUL_TILE_320x104", "CUBLASLT_MATMUL_TILE_312x64", "CUBLASLT_MATMUL_TILE_312x128", "CUBLASLT_MATMUL_TILE_304x64", @@ -12857,10 +12881,28 @@ sub warnHipOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_256x48", "CUBLASLT_MATMUL_TILE_256x40", "CUBLASLT_MATMUL_TILE_256x32", + "CUBLASLT_MATMUL_TILE_256x256", + "CUBLASLT_MATMUL_TILE_256x248", + "CUBLASLT_MATMUL_TILE_256x240", "CUBLASLT_MATMUL_TILE_256x24", + "CUBLASLT_MATMUL_TILE_256x232", + "CUBLASLT_MATMUL_TILE_256x224", + "CUBLASLT_MATMUL_TILE_256x216", + "CUBLASLT_MATMUL_TILE_256x208", + "CUBLASLT_MATMUL_TILE_256x200", "CUBLASLT_MATMUL_TILE_256x192", + "CUBLASLT_MATMUL_TILE_256x184", + "CUBLASLT_MATMUL_TILE_256x176", + "CUBLASLT_MATMUL_TILE_256x168", + "CUBLASLT_MATMUL_TILE_256x160", "CUBLASLT_MATMUL_TILE_256x16", + "CUBLASLT_MATMUL_TILE_256x152", + "CUBLASLT_MATMUL_TILE_256x144", + "CUBLASLT_MATMUL_TILE_256x136", "CUBLASLT_MATMUL_TILE_256x128", + "CUBLASLT_MATMUL_TILE_256x120", + "CUBLASLT_MATMUL_TILE_256x112", + "CUBLASLT_MATMUL_TILE_256x104", "CUBLASLT_MATMUL_TILE_24x768", "CUBLASLT_MATMUL_TILE_24x704", "CUBLASLT_MATMUL_TILE_24x640", @@ -14655,6 +14697,7 @@ sub warnRocOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_408x64", "CUBLASLT_MATMUL_TILE_400x64", "CUBLASLT_MATMUL_TILE_392x64", + "CUBLASLT_MATMUL_TILE_384x8", "CUBLASLT_MATMUL_TILE_384x64", "CUBLASLT_MATMUL_TILE_384x128", "CUBLASLT_MATMUL_TILE_376x64", @@ -14686,8 +14729,31 @@ sub warnRocOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_32x128", "CUBLASLT_MATMUL_TILE_328x64", "CUBLASLT_MATMUL_TILE_328x128", + "CUBLASLT_MATMUL_TILE_320x96", + "CUBLASLT_MATMUL_TILE_320x88", + "CUBLASLT_MATMUL_TILE_320x80", + "CUBLASLT_MATMUL_TILE_320x8", + "CUBLASLT_MATMUL_TILE_320x72", "CUBLASLT_MATMUL_TILE_320x64", + "CUBLASLT_MATMUL_TILE_320x56", + "CUBLASLT_MATMUL_TILE_320x48", + "CUBLASLT_MATMUL_TILE_320x40", + "CUBLASLT_MATMUL_TILE_320x32", + "CUBLASLT_MATMUL_TILE_320x24", + "CUBLASLT_MATMUL_TILE_320x200", + "CUBLASLT_MATMUL_TILE_320x192", + "CUBLASLT_MATMUL_TILE_320x184", + "CUBLASLT_MATMUL_TILE_320x176", + "CUBLASLT_MATMUL_TILE_320x168", + "CUBLASLT_MATMUL_TILE_320x160", + "CUBLASLT_MATMUL_TILE_320x16", + "CUBLASLT_MATMUL_TILE_320x152", + "CUBLASLT_MATMUL_TILE_320x144", + "CUBLASLT_MATMUL_TILE_320x136", "CUBLASLT_MATMUL_TILE_320x128", + "CUBLASLT_MATMUL_TILE_320x120", + "CUBLASLT_MATMUL_TILE_320x112", + "CUBLASLT_MATMUL_TILE_320x104", "CUBLASLT_MATMUL_TILE_312x64", "CUBLASLT_MATMUL_TILE_312x128", "CUBLASLT_MATMUL_TILE_304x64", @@ -14712,10 +14778,28 @@ sub warnRocOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_256x48", "CUBLASLT_MATMUL_TILE_256x40", "CUBLASLT_MATMUL_TILE_256x32", + "CUBLASLT_MATMUL_TILE_256x256", + "CUBLASLT_MATMUL_TILE_256x248", + "CUBLASLT_MATMUL_TILE_256x240", "CUBLASLT_MATMUL_TILE_256x24", + "CUBLASLT_MATMUL_TILE_256x232", + "CUBLASLT_MATMUL_TILE_256x224", + "CUBLASLT_MATMUL_TILE_256x216", + "CUBLASLT_MATMUL_TILE_256x208", + "CUBLASLT_MATMUL_TILE_256x200", "CUBLASLT_MATMUL_TILE_256x192", + "CUBLASLT_MATMUL_TILE_256x184", + "CUBLASLT_MATMUL_TILE_256x176", + "CUBLASLT_MATMUL_TILE_256x168", + "CUBLASLT_MATMUL_TILE_256x160", "CUBLASLT_MATMUL_TILE_256x16", + "CUBLASLT_MATMUL_TILE_256x152", + "CUBLASLT_MATMUL_TILE_256x144", + "CUBLASLT_MATMUL_TILE_256x136", "CUBLASLT_MATMUL_TILE_256x128", + "CUBLASLT_MATMUL_TILE_256x120", + "CUBLASLT_MATMUL_TILE_256x112", + "CUBLASLT_MATMUL_TILE_256x104", "CUBLASLT_MATMUL_TILE_24x768", "CUBLASLT_MATMUL_TILE_24x704", "CUBLASLT_MATMUL_TILE_24x640", diff --git a/docs/tables/CUBLAS_API_supported_by_HIP.md b/docs/tables/CUBLAS_API_supported_by_HIP.md index 7561d5c8..0076949a 100644 --- a/docs/tables/CUBLAS_API_supported_by_HIP.md +++ b/docs/tables/CUBLAS_API_supported_by_HIP.md @@ -526,10 +526,28 @@ |`CUBLASLT_MATMUL_TILE_24x640`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_24x704`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_24x768`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x104`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x112`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x120`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x128`|10.1| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x136`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x144`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x152`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x16`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x160`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x168`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x176`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x184`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x192`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x200`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x208`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x216`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x224`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x232`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x24`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x240`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x248`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x256`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x32`|12.1| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x40`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x48`|12.6| | | | | | | | | | @@ -554,8 +572,31 @@ |`CUBLASLT_MATMUL_TILE_304x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_312x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_312x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x104`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x112`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x120`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_320x128`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x136`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x144`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x152`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x16`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x160`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x168`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x176`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x184`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x192`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x200`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x24`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x32`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x40`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x48`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x56`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_320x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x72`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x8`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x80`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x88`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x96`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_328x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_328x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_32x128`|10.1| | | | | | | | | | @@ -587,6 +628,7 @@ |`CUBLASLT_MATMUL_TILE_376x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_384x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_384x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_384x8`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_392x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_400x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_408x64`|12.6| | | | | | | | | | diff --git a/docs/tables/CUBLAS_API_supported_by_HIP_and_ROC.md b/docs/tables/CUBLAS_API_supported_by_HIP_and_ROC.md index ae05ad0e..09ee7a09 100644 --- a/docs/tables/CUBLAS_API_supported_by_HIP_and_ROC.md +++ b/docs/tables/CUBLAS_API_supported_by_HIP_and_ROC.md @@ -526,10 +526,28 @@ |`CUBLASLT_MATMUL_TILE_24x640`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_24x704`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_24x768`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x104`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x112`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x120`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x128`|10.1| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x136`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x144`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x152`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x16`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x160`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x168`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x176`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x184`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x192`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x200`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x208`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x216`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x224`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x232`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x24`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x240`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x248`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x256`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x32`|12.1| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x40`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x48`|12.6| | | | | | | | | | | | | | | | @@ -554,8 +572,31 @@ |`CUBLASLT_MATMUL_TILE_304x64`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_312x128`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_312x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x104`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x112`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x120`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_320x128`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x136`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x144`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x152`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x16`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x160`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x168`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x176`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x184`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x192`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x200`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x24`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x32`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x40`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x48`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x56`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_320x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x72`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x8`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x80`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x88`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x96`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_328x128`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_328x64`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_32x128`|10.1| | | | | | | | | | | | | | | | @@ -587,6 +628,7 @@ |`CUBLASLT_MATMUL_TILE_376x64`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_384x128`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_384x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_384x8`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_392x64`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_400x64`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_408x64`|12.6| | | | | | | | | | | | | | | | diff --git a/docs/tables/CUBLAS_API_supported_by_ROC.md b/docs/tables/CUBLAS_API_supported_by_ROC.md index faf0f611..182d4f23 100644 --- a/docs/tables/CUBLAS_API_supported_by_ROC.md +++ b/docs/tables/CUBLAS_API_supported_by_ROC.md @@ -526,10 +526,28 @@ |`CUBLASLT_MATMUL_TILE_24x640`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_24x704`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_24x768`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x104`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x112`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x120`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x128`|10.1| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x136`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x144`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x152`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x16`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x160`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x168`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x176`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x184`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x192`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x200`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x208`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x216`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x224`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x232`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x24`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x240`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x248`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_256x256`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x32`|12.1| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x40`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_256x48`|12.6| | | | | | | | | | @@ -554,8 +572,31 @@ |`CUBLASLT_MATMUL_TILE_304x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_312x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_312x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x104`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x112`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x120`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_320x128`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x136`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x144`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x152`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x16`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x160`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x168`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x176`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x184`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x192`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x200`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x24`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x32`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x40`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x48`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x56`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_320x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x72`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x8`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x80`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x88`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_320x96`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_328x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_328x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_32x128`|10.1| | | | | | | | | | @@ -587,6 +628,7 @@ |`CUBLASLT_MATMUL_TILE_376x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_384x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_384x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_384x8`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_392x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_400x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_408x64`|12.6| | | | | | | | | | diff --git a/src/CUDA2HIP_BLAS_API_types.cpp b/src/CUDA2HIP_BLAS_API_types.cpp index aaaa92ba..7ef3c214 100644 --- a/src/CUDA2HIP_BLAS_API_types.cpp +++ b/src/CUDA2HIP_BLAS_API_types.cpp @@ -651,54 +651,96 @@ const std::map CUDA_BLAS_TYPE_NAME_MAP { {"CUBLASLT_MATMUL_TILE_128x496", {"HIPBLASLT_MATMUL_TILE_128x496", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, {"CUBLASLT_MATMUL_TILE_128x504", {"HIPBLASLT_MATMUL_TILE_128x504", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, {"CUBLASLT_MATMUL_TILE_128x512", {"HIPBLASLT_MATMUL_TILE_128x512", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x8", {"CUBLASLT_MATMUL_TILE_192x8", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x16", {"CUBLASLT_MATMUL_TILE_192x16", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x24", {"CUBLASLT_MATMUL_TILE_192x24", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x32", {"CUBLASLT_MATMUL_TILE_192x32", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x40", {"CUBLASLT_MATMUL_TILE_192x40", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x48", {"CUBLASLT_MATMUL_TILE_192x48", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x56", {"CUBLASLT_MATMUL_TILE_192x56", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x72", {"CUBLASLT_MATMUL_TILE_192x72", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x80", {"CUBLASLT_MATMUL_TILE_192x80", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x88", {"CUBLASLT_MATMUL_TILE_192x88", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x96", {"CUBLASLT_MATMUL_TILE_192x96", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x104", {"CUBLASLT_MATMUL_TILE_192x104", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x112", {"CUBLASLT_MATMUL_TILE_192x112", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x120", {"CUBLASLT_MATMUL_TILE_192x120", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x136", {"CUBLASLT_MATMUL_TILE_192x136", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x144", {"CUBLASLT_MATMUL_TILE_192x144", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x152", {"CUBLASLT_MATMUL_TILE_192x152", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x160", {"CUBLASLT_MATMUL_TILE_192x160", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x168", {"CUBLASLT_MATMUL_TILE_192x168", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x176", {"CUBLASLT_MATMUL_TILE_192x176", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x184", {"CUBLASLT_MATMUL_TILE_192x184", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x200", {"CUBLASLT_MATMUL_TILE_192x200", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x208", {"CUBLASLT_MATMUL_TILE_192x208", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x216", {"CUBLASLT_MATMUL_TILE_192x216", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x224", {"CUBLASLT_MATMUL_TILE_192x224", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x232", {"CUBLASLT_MATMUL_TILE_192x232", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x240", {"CUBLASLT_MATMUL_TILE_192x240", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x248", {"CUBLASLT_MATMUL_TILE_192x248", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x264", {"CUBLASLT_MATMUL_TILE_192x264", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x272", {"CUBLASLT_MATMUL_TILE_192x272", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x280", {"CUBLASLT_MATMUL_TILE_192x280", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x288", {"CUBLASLT_MATMUL_TILE_192x288", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x296", {"CUBLASLT_MATMUL_TILE_192x296", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x304", {"CUBLASLT_MATMUL_TILE_192x304", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x312", {"CUBLASLT_MATMUL_TILE_192x312", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x320", {"CUBLASLT_MATMUL_TILE_192x320", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x328", {"CUBLASLT_MATMUL_TILE_192x328", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_192x336", {"CUBLASLT_MATMUL_TILE_192x336", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_256x8", {"CUBLASLT_MATMUL_TILE_256x8", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_256x16", {"CUBLASLT_MATMUL_TILE_256x16", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_256x24", {"CUBLASLT_MATMUL_TILE_256x24", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_256x40", {"CUBLASLT_MATMUL_TILE_256x40", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_256x48", {"CUBLASLT_MATMUL_TILE_256x48", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_256x56", {"CUBLASLT_MATMUL_TILE_256x56", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_256x72", {"CUBLASLT_MATMUL_TILE_256x72", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_256x80", {"CUBLASLT_MATMUL_TILE_256x80", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_256x88", {"CUBLASLT_MATMUL_TILE_256x88", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, - {"CUBLASLT_MATMUL_TILE_256x96", {"CUBLASLT_MATMUL_TILE_256x96", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x8", {"HIPBLASLT_MATMUL_TILE_192x8", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x16", {"HIPBLASLT_MATMUL_TILE_192x16", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x24", {"HIPBLASLT_MATMUL_TILE_192x24", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x32", {"HIPBLASLT_MATMUL_TILE_192x32", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x40", {"HIPBLASLT_MATMUL_TILE_192x40", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x48", {"HIPBLASLT_MATMUL_TILE_192x48", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x56", {"HIPBLASLT_MATMUL_TILE_192x56", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x72", {"HIPBLASLT_MATMUL_TILE_192x72", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x80", {"HIPBLASLT_MATMUL_TILE_192x80", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x88", {"HIPBLASLT_MATMUL_TILE_192x88", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x96", {"HIPBLASLT_MATMUL_TILE_192x96", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x104", {"HIPBLASLT_MATMUL_TILE_192x104", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x112", {"HIPBLASLT_MATMUL_TILE_192x112", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x120", {"HIPBLASLT_MATMUL_TILE_192x120", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x136", {"HIPBLASLT_MATMUL_TILE_192x136", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x144", {"HIPBLASLT_MATMUL_TILE_192x144", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x152", {"HIPBLASLT_MATMUL_TILE_192x152", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x160", {"HIPBLASLT_MATMUL_TILE_192x160", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x168", {"HIPBLASLT_MATMUL_TILE_192x168", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x176", {"HIPBLASLT_MATMUL_TILE_192x176", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x184", {"HIPBLASLT_MATMUL_TILE_192x184", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x200", {"HIPBLASLT_MATMUL_TILE_192x200", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x208", {"HIPBLASLT_MATMUL_TILE_192x208", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x216", {"HIPBLASLT_MATMUL_TILE_192x216", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x224", {"HIPBLASLT_MATMUL_TILE_192x224", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x232", {"HIPBLASLT_MATMUL_TILE_192x232", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x240", {"HIPBLASLT_MATMUL_TILE_192x240", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x248", {"HIPBLASLT_MATMUL_TILE_192x248", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x264", {"HIPBLASLT_MATMUL_TILE_192x264", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x272", {"HIPBLASLT_MATMUL_TILE_192x272", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x280", {"HIPBLASLT_MATMUL_TILE_192x280", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x288", {"HIPBLASLT_MATMUL_TILE_192x288", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x296", {"HIPBLASLT_MATMUL_TILE_192x296", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x304", {"HIPBLASLT_MATMUL_TILE_192x304", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x312", {"HIPBLASLT_MATMUL_TILE_192x312", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x320", {"HIPBLASLT_MATMUL_TILE_192x320", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x328", {"HIPBLASLT_MATMUL_TILE_192x328", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_192x336", {"HIPBLASLT_MATMUL_TILE_192x336", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x8", {"HIPBLASLT_MATMUL_TILE_256x8", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x16", {"HIPBLASLT_MATMUL_TILE_256x16", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x24", {"HIPBLASLT_MATMUL_TILE_256x24", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x40", {"HIPBLASLT_MATMUL_TILE_256x40", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x48", {"HIPBLASLT_MATMUL_TILE_256x48", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x56", {"HIPBLASLT_MATMUL_TILE_256x56", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x72", {"HIPBLASLT_MATMUL_TILE_256x72", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x80", {"HIPBLASLT_MATMUL_TILE_256x80", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x88", {"HIPBLASLT_MATMUL_TILE_256x88", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x96", {"HIPBLASLT_MATMUL_TILE_256x96", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x104", {"HIPBLASLT_MATMUL_TILE_256x104", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x112", {"HIPBLASLT_MATMUL_TILE_256x112", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x120", {"HIPBLASLT_MATMUL_TILE_256x120", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x136", {"HIPBLASLT_MATMUL_TILE_256x136", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x144", {"HIPBLASLT_MATMUL_TILE_256x144", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x152", {"HIPBLASLT_MATMUL_TILE_256x152", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x160", {"HIPBLASLT_MATMUL_TILE_256x160", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x168", {"HIPBLASLT_MATMUL_TILE_256x168", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x176", {"HIPBLASLT_MATMUL_TILE_256x176", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x184", {"HIPBLASLT_MATMUL_TILE_256x184", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x200", {"HIPBLASLT_MATMUL_TILE_256x200", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x208", {"HIPBLASLT_MATMUL_TILE_256x208", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x216", {"HIPBLASLT_MATMUL_TILE_256x216", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x224", {"HIPBLASLT_MATMUL_TILE_256x224", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x232", {"HIPBLASLT_MATMUL_TILE_256x232", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x240", {"HIPBLASLT_MATMUL_TILE_256x240", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x248", {"HIPBLASLT_MATMUL_TILE_256x248", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_256x256", {"HIPBLASLT_MATMUL_TILE_256x256", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x8", {"HIPBLASLT_MATMUL_TILE_320x8", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x16", {"HIPBLASLT_MATMUL_TILE_320x16", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x24", {"HIPBLASLT_MATMUL_TILE_320x24", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x32", {"HIPBLASLT_MATMUL_TILE_320x32", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x40", {"HIPBLASLT_MATMUL_TILE_320x40", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x48", {"HIPBLASLT_MATMUL_TILE_320x48", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x56", {"HIPBLASLT_MATMUL_TILE_320x56", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x72", {"HIPBLASLT_MATMUL_TILE_320x72", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x80", {"HIPBLASLT_MATMUL_TILE_320x80", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x88", {"HIPBLASLT_MATMUL_TILE_320x88", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x96", {"HIPBLASLT_MATMUL_TILE_320x96", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x104", {"HIPBLASLT_MATMUL_TILE_320x104", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x112", {"HIPBLASLT_MATMUL_TILE_320x112", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x120", {"HIPBLASLT_MATMUL_TILE_320x120", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x136", {"HIPBLASLT_MATMUL_TILE_320x136", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x144", {"HIPBLASLT_MATMUL_TILE_320x144", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x152", {"HIPBLASLT_MATMUL_TILE_320x152", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x160", {"HIPBLASLT_MATMUL_TILE_320x160", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x168", {"HIPBLASLT_MATMUL_TILE_320x168", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x176", {"HIPBLASLT_MATMUL_TILE_320x176", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x184", {"HIPBLASLT_MATMUL_TILE_320x184", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x192", {"HIPBLASLT_MATMUL_TILE_320x192", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_320x200", {"HIPBLASLT_MATMUL_TILE_320x200", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_384x8", {"HIPBLASLT_MATMUL_TILE_384x8", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, {"CUBLASLT_MATMUL_TILE_END", {"HIPBLASLT_MATMUL_TILE_END", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, {"cublasLtMatmulStages_t", {"hipblasLtMatmulStages_t", "", CONV_TYPE, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, {"CUBLASLT_MATMUL_STAGES_UNDEFINED", {"HIPBLASLT_MATMUL_STAGES_UNDEFINED", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, @@ -1815,6 +1857,48 @@ const std::map CUDA_BLAS_TYPE_NAME_VER_MAP { {"CUBLASLT_MATMUL_TILE_256x80", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 {"CUBLASLT_MATMUL_TILE_256x88", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 {"CUBLASLT_MATMUL_TILE_256x96", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x104", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x112", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x120", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x136", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x144", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x152", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x160", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x168", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x176", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x184", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x200", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x208", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x216", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x224", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x232", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x240", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x248", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_256x256", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x8", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x16", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x24", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x32", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x40", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x48", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x56", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x72", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x80", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x88", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x96", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x104", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x112", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x120", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x136", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x144", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x152", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x160", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x168", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x176", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x184", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x192", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_320x200", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 + {"CUBLASLT_MATMUL_TILE_384x8", {CUDA_126, CUDA_0, CUDA_0 }}, // A: CUDA_VERSION 12062, CUBLAS_VERSION 120603, CUBLAS_VER_MAJOR 12 CUBLAS_VER_MINOR 6 CUBLAS_VER_PATCH 3 }; const std::map HIP_BLAS_TYPE_NAME_VER_MAP {