diff --git a/bin/hipify-perl b/bin/hipify-perl index 68dbcf16..ed9d63ce 100755 --- a/bin/hipify-perl +++ b/bin/hipify-perl @@ -12583,6 +12583,11 @@ sub warnHipOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_80x256", "CUBLASLT_MATMUL_TILE_80x192", "CUBLASLT_MATMUL_TILE_80x128", + "CUBLASLT_MATMUL_TILE_768x64", + "CUBLASLT_MATMUL_TILE_760x64", + "CUBLASLT_MATMUL_TILE_752x64", + "CUBLASLT_MATMUL_TILE_744x64", + "CUBLASLT_MATMUL_TILE_736x64", "CUBLASLT_MATMUL_TILE_72x640", "CUBLASLT_MATMUL_TILE_72x64", "CUBLASLT_MATMUL_TILE_72x576", @@ -12593,6 +12598,16 @@ sub warnHipOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_72x256", "CUBLASLT_MATMUL_TILE_72x192", "CUBLASLT_MATMUL_TILE_72x128", + "CUBLASLT_MATMUL_TILE_728x64", + "CUBLASLT_MATMUL_TILE_720x64", + "CUBLASLT_MATMUL_TILE_712x64", + "CUBLASLT_MATMUL_TILE_704x64", + "CUBLASLT_MATMUL_TILE_696x64", + "CUBLASLT_MATMUL_TILE_688x64", + "CUBLASLT_MATMUL_TILE_680x64", + "CUBLASLT_MATMUL_TILE_672x64", + "CUBLASLT_MATMUL_TILE_664x64", + "CUBLASLT_MATMUL_TILE_656x64", "CUBLASLT_MATMUL_TILE_64x96", "CUBLASLT_MATMUL_TILE_64x8", "CUBLASLT_MATMUL_TILE_64x768", @@ -12608,6 +12623,16 @@ sub warnHipOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_64x256", "CUBLASLT_MATMUL_TILE_64x192", "CUBLASLT_MATMUL_TILE_64x128", + "CUBLASLT_MATMUL_TILE_648x64", + "CUBLASLT_MATMUL_TILE_640x64", + "CUBLASLT_MATMUL_TILE_632x64", + "CUBLASLT_MATMUL_TILE_624x64", + "CUBLASLT_MATMUL_TILE_616x64", + "CUBLASLT_MATMUL_TILE_608x64", + "CUBLASLT_MATMUL_TILE_600x64", + "CUBLASLT_MATMUL_TILE_592x64", + "CUBLASLT_MATMUL_TILE_584x64", + "CUBLASLT_MATMUL_TILE_576x64", "CUBLASLT_MATMUL_TILE_56x768", "CUBLASLT_MATMUL_TILE_56x704", "CUBLASLT_MATMUL_TILE_56x640", @@ -12619,7 +12644,16 @@ sub warnHipOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_56x256", "CUBLASLT_MATMUL_TILE_56x192", "CUBLASLT_MATMUL_TILE_56x128", + "CUBLASLT_MATMUL_TILE_568x64", + "CUBLASLT_MATMUL_TILE_560x64", + "CUBLASLT_MATMUL_TILE_552x64", + "CUBLASLT_MATMUL_TILE_544x64", + "CUBLASLT_MATMUL_TILE_536x64", + "CUBLASLT_MATMUL_TILE_528x64", + "CUBLASLT_MATMUL_TILE_520x64", "CUBLASLT_MATMUL_TILE_512x64", + "CUBLASLT_MATMUL_TILE_504x64", + "CUBLASLT_MATMUL_TILE_496x64", "CUBLASLT_MATMUL_TILE_48x768", "CUBLASLT_MATMUL_TILE_48x704", "CUBLASLT_MATMUL_TILE_48x640", @@ -12632,6 +12666,16 @@ sub warnHipOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_48x256", "CUBLASLT_MATMUL_TILE_48x192", "CUBLASLT_MATMUL_TILE_48x128", + "CUBLASLT_MATMUL_TILE_488x64", + "CUBLASLT_MATMUL_TILE_480x64", + "CUBLASLT_MATMUL_TILE_472x64", + "CUBLASLT_MATMUL_TILE_464x64", + "CUBLASLT_MATMUL_TILE_456x64", + "CUBLASLT_MATMUL_TILE_448x64", + "CUBLASLT_MATMUL_TILE_440x64", + "CUBLASLT_MATMUL_TILE_432x64", + "CUBLASLT_MATMUL_TILE_424x64", + "CUBLASLT_MATMUL_TILE_416x64", "CUBLASLT_MATMUL_TILE_40x768", "CUBLASLT_MATMUL_TILE_40x704", "CUBLASLT_MATMUL_TILE_40x640", @@ -12644,6 +12688,8 @@ sub warnHipOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_40x256", "CUBLASLT_MATMUL_TILE_40x192", "CUBLASLT_MATMUL_TILE_40x128", + "CUBLASLT_MATMUL_TILE_408x64", + "CUBLASLT_MATMUL_TILE_400x64", "CUBLASLT_MATMUL_TILE_392x64", "CUBLASLT_MATMUL_TILE_384x64", "CUBLASLT_MATMUL_TILE_384x128", @@ -14236,6 +14282,11 @@ sub warnRocOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_80x256", "CUBLASLT_MATMUL_TILE_80x192", "CUBLASLT_MATMUL_TILE_80x128", + "CUBLASLT_MATMUL_TILE_768x64", + "CUBLASLT_MATMUL_TILE_760x64", + "CUBLASLT_MATMUL_TILE_752x64", + "CUBLASLT_MATMUL_TILE_744x64", + "CUBLASLT_MATMUL_TILE_736x64", "CUBLASLT_MATMUL_TILE_72x640", "CUBLASLT_MATMUL_TILE_72x64", "CUBLASLT_MATMUL_TILE_72x576", @@ -14246,6 +14297,16 @@ sub warnRocOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_72x256", "CUBLASLT_MATMUL_TILE_72x192", "CUBLASLT_MATMUL_TILE_72x128", + "CUBLASLT_MATMUL_TILE_728x64", + "CUBLASLT_MATMUL_TILE_720x64", + "CUBLASLT_MATMUL_TILE_712x64", + "CUBLASLT_MATMUL_TILE_704x64", + "CUBLASLT_MATMUL_TILE_696x64", + "CUBLASLT_MATMUL_TILE_688x64", + "CUBLASLT_MATMUL_TILE_680x64", + "CUBLASLT_MATMUL_TILE_672x64", + "CUBLASLT_MATMUL_TILE_664x64", + "CUBLASLT_MATMUL_TILE_656x64", "CUBLASLT_MATMUL_TILE_64x96", "CUBLASLT_MATMUL_TILE_64x8", "CUBLASLT_MATMUL_TILE_64x768", @@ -14261,6 +14322,16 @@ sub warnRocOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_64x256", "CUBLASLT_MATMUL_TILE_64x192", "CUBLASLT_MATMUL_TILE_64x128", + "CUBLASLT_MATMUL_TILE_648x64", + "CUBLASLT_MATMUL_TILE_640x64", + "CUBLASLT_MATMUL_TILE_632x64", + "CUBLASLT_MATMUL_TILE_624x64", + "CUBLASLT_MATMUL_TILE_616x64", + "CUBLASLT_MATMUL_TILE_608x64", + "CUBLASLT_MATMUL_TILE_600x64", + "CUBLASLT_MATMUL_TILE_592x64", + "CUBLASLT_MATMUL_TILE_584x64", + "CUBLASLT_MATMUL_TILE_576x64", "CUBLASLT_MATMUL_TILE_56x768", "CUBLASLT_MATMUL_TILE_56x704", "CUBLASLT_MATMUL_TILE_56x640", @@ -14272,7 +14343,16 @@ sub warnRocOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_56x256", "CUBLASLT_MATMUL_TILE_56x192", "CUBLASLT_MATMUL_TILE_56x128", + "CUBLASLT_MATMUL_TILE_568x64", + "CUBLASLT_MATMUL_TILE_560x64", + "CUBLASLT_MATMUL_TILE_552x64", + "CUBLASLT_MATMUL_TILE_544x64", + "CUBLASLT_MATMUL_TILE_536x64", + "CUBLASLT_MATMUL_TILE_528x64", + "CUBLASLT_MATMUL_TILE_520x64", "CUBLASLT_MATMUL_TILE_512x64", + "CUBLASLT_MATMUL_TILE_504x64", + "CUBLASLT_MATMUL_TILE_496x64", "CUBLASLT_MATMUL_TILE_48x768", "CUBLASLT_MATMUL_TILE_48x704", "CUBLASLT_MATMUL_TILE_48x640", @@ -14285,6 +14365,16 @@ sub warnRocOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_48x256", "CUBLASLT_MATMUL_TILE_48x192", "CUBLASLT_MATMUL_TILE_48x128", + "CUBLASLT_MATMUL_TILE_488x64", + "CUBLASLT_MATMUL_TILE_480x64", + "CUBLASLT_MATMUL_TILE_472x64", + "CUBLASLT_MATMUL_TILE_464x64", + "CUBLASLT_MATMUL_TILE_456x64", + "CUBLASLT_MATMUL_TILE_448x64", + "CUBLASLT_MATMUL_TILE_440x64", + "CUBLASLT_MATMUL_TILE_432x64", + "CUBLASLT_MATMUL_TILE_424x64", + "CUBLASLT_MATMUL_TILE_416x64", "CUBLASLT_MATMUL_TILE_40x768", "CUBLASLT_MATMUL_TILE_40x704", "CUBLASLT_MATMUL_TILE_40x640", @@ -14297,6 +14387,8 @@ sub warnRocOnlyUnsupportedFunctions { "CUBLASLT_MATMUL_TILE_40x256", "CUBLASLT_MATMUL_TILE_40x192", "CUBLASLT_MATMUL_TILE_40x128", + "CUBLASLT_MATMUL_TILE_408x64", + "CUBLASLT_MATMUL_TILE_400x64", "CUBLASLT_MATMUL_TILE_392x64", "CUBLASLT_MATMUL_TILE_384x64", "CUBLASLT_MATMUL_TILE_384x128", diff --git a/docs/tables/CUBLAS_API_supported_by_HIP.md b/docs/tables/CUBLAS_API_supported_by_HIP.md index 8f6066c5..1bb0bdc8 100644 --- a/docs/tables/CUBLAS_API_supported_by_HIP.md +++ b/docs/tables/CUBLAS_API_supported_by_HIP.md @@ -485,6 +485,8 @@ |`CUBLASLT_MATMUL_TILE_384x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_384x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_392x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_400x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_408x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_40x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_40x192`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_40x256`|12.6| | | | | | | | | | @@ -497,6 +499,16 @@ |`CUBLASLT_MATMUL_TILE_40x640`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_40x704`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_40x768`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_416x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_424x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_432x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_440x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_448x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_456x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_464x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_472x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_480x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_488x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_48x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_48x192`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_48x256`|12.6| | | | | | | | | | @@ -509,7 +521,16 @@ |`CUBLASLT_MATMUL_TILE_48x640`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_48x704`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_48x768`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_496x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_504x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_512x64`|10.1| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_520x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_528x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_536x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_544x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_552x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_560x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_568x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_56x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_56x192`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_56x256`|12.6| | | | | | | | | | @@ -521,6 +542,16 @@ |`CUBLASLT_MATMUL_TILE_56x640`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_56x704`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_56x768`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_576x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_584x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_592x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_600x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_608x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_616x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_624x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_632x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_640x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_648x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_64x128`|10.1| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_64x192`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_64x256`|10.1| | | | | | | | | | @@ -536,6 +567,16 @@ |`CUBLASLT_MATMUL_TILE_64x768`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_64x8`|10.1| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_64x96`|11.3| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_656x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_664x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_672x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_680x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_688x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_696x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_704x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_712x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_720x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_728x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_72x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_72x192`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_72x256`|12.6| | | | | | | | | | @@ -546,6 +587,11 @@ |`CUBLASLT_MATMUL_TILE_72x576`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_72x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_72x640`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_736x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_744x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_752x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_760x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_768x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_80x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_80x192`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_80x256`|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 55f789f4..824abff6 100644 --- a/docs/tables/CUBLAS_API_supported_by_HIP_and_ROC.md +++ b/docs/tables/CUBLAS_API_supported_by_HIP_and_ROC.md @@ -485,6 +485,8 @@ |`CUBLASLT_MATMUL_TILE_384x128`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_384x64`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_392x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_400x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_408x64`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_40x128`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_40x192`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_40x256`|12.6| | | | | | | | | | | | | | | | @@ -497,6 +499,16 @@ |`CUBLASLT_MATMUL_TILE_40x640`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_40x704`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_40x768`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_416x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_424x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_432x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_440x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_448x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_456x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_464x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_472x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_480x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_488x64`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_48x128`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_48x192`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_48x256`|12.6| | | | | | | | | | | | | | | | @@ -509,7 +521,16 @@ |`CUBLASLT_MATMUL_TILE_48x640`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_48x704`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_48x768`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_496x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_504x64`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_512x64`|10.1| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_520x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_528x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_536x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_544x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_552x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_560x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_568x64`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_56x128`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_56x192`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_56x256`|12.6| | | | | | | | | | | | | | | | @@ -521,6 +542,16 @@ |`CUBLASLT_MATMUL_TILE_56x640`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_56x704`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_56x768`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_576x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_584x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_592x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_600x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_608x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_616x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_624x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_632x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_640x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_648x64`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_64x128`|10.1| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_64x192`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_64x256`|10.1| | | | | | | | | | | | | | | | @@ -536,6 +567,16 @@ |`CUBLASLT_MATMUL_TILE_64x768`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_64x8`|10.1| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_64x96`|11.3| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_656x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_664x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_672x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_680x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_688x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_696x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_704x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_712x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_720x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_728x64`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_72x128`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_72x192`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_72x256`|12.6| | | | | | | | | | | | | | | | @@ -546,6 +587,11 @@ |`CUBLASLT_MATMUL_TILE_72x576`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_72x64`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_72x640`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_736x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_744x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_752x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_760x64`|12.6| | | | | | | | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_768x64`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_80x128`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_80x192`|12.6| | | | | | | | | | | | | | | | |`CUBLASLT_MATMUL_TILE_80x256`|12.6| | | | | | | | | | | | | | | | diff --git a/docs/tables/CUBLAS_API_supported_by_ROC.md b/docs/tables/CUBLAS_API_supported_by_ROC.md index 578bc19d..6d135850 100644 --- a/docs/tables/CUBLAS_API_supported_by_ROC.md +++ b/docs/tables/CUBLAS_API_supported_by_ROC.md @@ -485,6 +485,8 @@ |`CUBLASLT_MATMUL_TILE_384x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_384x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_392x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_400x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_408x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_40x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_40x192`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_40x256`|12.6| | | | | | | | | | @@ -497,6 +499,16 @@ |`CUBLASLT_MATMUL_TILE_40x640`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_40x704`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_40x768`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_416x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_424x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_432x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_440x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_448x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_456x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_464x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_472x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_480x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_488x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_48x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_48x192`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_48x256`|12.6| | | | | | | | | | @@ -509,7 +521,16 @@ |`CUBLASLT_MATMUL_TILE_48x640`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_48x704`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_48x768`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_496x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_504x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_512x64`|10.1| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_520x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_528x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_536x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_544x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_552x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_560x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_568x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_56x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_56x192`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_56x256`|12.6| | | | | | | | | | @@ -521,6 +542,16 @@ |`CUBLASLT_MATMUL_TILE_56x640`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_56x704`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_56x768`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_576x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_584x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_592x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_600x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_608x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_616x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_624x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_632x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_640x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_648x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_64x128`|10.1| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_64x192`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_64x256`|10.1| | | | | | | | | | @@ -536,6 +567,16 @@ |`CUBLASLT_MATMUL_TILE_64x768`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_64x8`|10.1| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_64x96`|11.3| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_656x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_664x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_672x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_680x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_688x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_696x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_704x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_712x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_720x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_728x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_72x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_72x192`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_72x256`|12.6| | | | | | | | | | @@ -546,6 +587,11 @@ |`CUBLASLT_MATMUL_TILE_72x576`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_72x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_72x640`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_736x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_744x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_752x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_760x64`|12.6| | | | | | | | | | +|`CUBLASLT_MATMUL_TILE_768x64`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_80x128`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_80x192`|12.6| | | | | | | | | | |`CUBLASLT_MATMUL_TILE_80x256`|12.6| | | | | | | | | | diff --git a/src/CUDA2HIP_BLAS_API_types.cpp b/src/CUDA2HIP_BLAS_API_types.cpp index 747a27c0..b4c7aab3 100644 --- a/src/CUDA2HIP_BLAS_API_types.cpp +++ b/src/CUDA2HIP_BLAS_API_types.cpp @@ -469,6 +469,52 @@ const std::map CUDA_BLAS_TYPE_NAME_MAP { {"CUBLASLT_MATMUL_TILE_384x64", {"HIPBLASLT_MATMUL_TILE_384x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, {"CUBLASLT_MATMUL_TILE_384x128", {"HIPBLASLT_MATMUL_TILE_384x128", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, {"CUBLASLT_MATMUL_TILE_392x64", {"HIPBLASLT_MATMUL_TILE_392x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_400x64", {"HIPBLASLT_MATMUL_TILE_400x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_408x64", {"HIPBLASLT_MATMUL_TILE_408x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_416x64", {"HIPBLASLT_MATMUL_TILE_416x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_424x64", {"HIPBLASLT_MATMUL_TILE_424x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_432x64", {"HIPBLASLT_MATMUL_TILE_432x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_440x64", {"HIPBLASLT_MATMUL_TILE_440x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_448x64", {"HIPBLASLT_MATMUL_TILE_448x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_456x64", {"HIPBLASLT_MATMUL_TILE_456x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_464x64", {"HIPBLASLT_MATMUL_TILE_464x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_472x64", {"HIPBLASLT_MATMUL_TILE_472x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_480x64", {"HIPBLASLT_MATMUL_TILE_480x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_488x64", {"HIPBLASLT_MATMUL_TILE_488x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_496x64", {"HIPBLASLT_MATMUL_TILE_496x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_504x64", {"HIPBLASLT_MATMUL_TILE_504x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_520x64", {"HIPBLASLT_MATMUL_TILE_520x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_528x64", {"HIPBLASLT_MATMUL_TILE_528x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_536x64", {"HIPBLASLT_MATMUL_TILE_536x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_544x64", {"HIPBLASLT_MATMUL_TILE_544x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_552x64", {"HIPBLASLT_MATMUL_TILE_552x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_560x64", {"HIPBLASLT_MATMUL_TILE_560x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_568x64", {"HIPBLASLT_MATMUL_TILE_568x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_576x64", {"HIPBLASLT_MATMUL_TILE_576x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_584x64", {"HIPBLASLT_MATMUL_TILE_584x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_592x64", {"HIPBLASLT_MATMUL_TILE_592x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_600x64", {"HIPBLASLT_MATMUL_TILE_600x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_608x64", {"HIPBLASLT_MATMUL_TILE_608x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_616x64", {"HIPBLASLT_MATMUL_TILE_616x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_624x64", {"HIPBLASLT_MATMUL_TILE_624x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_632x64", {"HIPBLASLT_MATMUL_TILE_632x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_640x64", {"HIPBLASLT_MATMUL_TILE_640x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_648x64", {"HIPBLASLT_MATMUL_TILE_648x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_656x64", {"HIPBLASLT_MATMUL_TILE_656x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_664x64", {"HIPBLASLT_MATMUL_TILE_664x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_672x64", {"HIPBLASLT_MATMUL_TILE_672x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_680x64", {"HIPBLASLT_MATMUL_TILE_680x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_688x64", {"HIPBLASLT_MATMUL_TILE_688x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_696x64", {"HIPBLASLT_MATMUL_TILE_696x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_704x64", {"HIPBLASLT_MATMUL_TILE_704x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_712x64", {"HIPBLASLT_MATMUL_TILE_712x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_720x64", {"HIPBLASLT_MATMUL_TILE_720x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_728x64", {"HIPBLASLT_MATMUL_TILE_728x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_736x64", {"HIPBLASLT_MATMUL_TILE_736x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_744x64", {"HIPBLASLT_MATMUL_TILE_744x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_752x64", {"HIPBLASLT_MATMUL_TILE_752x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_760x64", {"HIPBLASLT_MATMUL_TILE_760x64", "", CONV_NUMERIC_LITERAL, API_BLAS, SEC::BLAS_LT_DATA_TYPES, UNSUPPORTED}}, + {"CUBLASLT_MATMUL_TILE_768x64", {"HIPBLASLT_MATMUL_TILE_768x64", "", 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}}, @@ -1355,6 +1401,52 @@ const std::map CUDA_BLAS_TYPE_NAME_VER_MAP { {"CUBLASLT_MATMUL_TILE_384x64", {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_384x128", {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_392x64", {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_400x64", {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_408x64", {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_416x64", {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_424x64", {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_432x64", {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_440x64", {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_448x64", {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_456x64", {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_464x64", {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_472x64", {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_480x64", {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_488x64", {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_496x64", {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_504x64", {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_520x64", {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_528x64", {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_536x64", {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_544x64", {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_552x64", {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_560x64", {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_568x64", {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_576x64", {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_584x64", {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_592x64", {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_600x64", {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_608x64", {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_616x64", {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_624x64", {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_632x64", {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_640x64", {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_648x64", {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_656x64", {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_664x64", {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_672x64", {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_680x64", {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_688x64", {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_696x64", {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_704x64", {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_712x64", {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_720x64", {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_728x64", {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_736x64", {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_744x64", {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_752x64", {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_760x64", {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_768x64", {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 {