derek.holloway
7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
136 lines
4.1 KiB
YAML
136 lines
4.1 KiB
YAML
GlobalParameters:
|
|
MinimumRequiredVersion: 4.4.0
|
|
PrintLevel: 1
|
|
ForceRedoBenchmarkProblems: True
|
|
ForceRedoLibraryLogic: True
|
|
ForceRedoLibraryClient: True
|
|
CMakeBuildType: Release
|
|
EnqueuesPerSync: 1
|
|
SyncsPerBenchmark: 1
|
|
LibraryPrintDebug: False
|
|
NumElementsToValidate: 1000
|
|
ValidationMaxToPrint: 4
|
|
ValidationPrintValids: False
|
|
ShortNames: False
|
|
MergeFiles: True
|
|
Platform: 0
|
|
Device: 0
|
|
KernelTime: True
|
|
PinClocks: True
|
|
SleepPercent: 200
|
|
DataInitTypeBeta : 0
|
|
|
|
BenchmarkProblems:
|
|
########################################
|
|
# NT - standard
|
|
########################################
|
|
-
|
|
- # ProblemType
|
|
OperationType: GEMM
|
|
DataType: d
|
|
TransposeA: False
|
|
TransposeB: True
|
|
UseBeta: True
|
|
Batched: True
|
|
|
|
- # BenchmarkProblemSizeGroup - Standard
|
|
InitialSolutionParameters:
|
|
BenchmarkCommonParameters:
|
|
- KernelLanguage: ["Assembly"]
|
|
- EdgeType: ["ShiftPtr"]
|
|
- LoopTail: [True]
|
|
ForkParameters:
|
|
- PrefetchLocalRead: [True,False]
|
|
- PrefetchGlobalRead: [True]
|
|
- ThreadTile:
|
|
- [ 6, 4 ]
|
|
- [ 8, 4 ]
|
|
- [ 6, 6 ]
|
|
- WorkGroup:
|
|
- [ 16, 16, 1 ]
|
|
- [ 16, 32, 1 ]
|
|
- WorkGroupMapping: [8]
|
|
- GlobalSplitU: [1]
|
|
- DepthU: [ 8 ]
|
|
- VectorWidth: [2]
|
|
- FractionalLoad : [1]
|
|
BenchmarkForkParameters:
|
|
JoinParameters:
|
|
- MacroTile
|
|
- GlobalSplitU
|
|
BenchmarkJoinParameters:
|
|
BenchmarkFinalParameters:
|
|
- ProblemSizes:
|
|
- Exact: [ 16, 16, 1, 16 ]
|
|
- Exact: [ 32, 32, 1, 32 ]
|
|
- Exact: [ 64, 64, 1, 64 ]
|
|
- Exact: [ 128, 128, 1, 128 ]
|
|
- Exact: [ 256, 256, 1, 256 ]
|
|
- Exact: [ 512, 512, 1, 512 ]
|
|
- Exact: [ 1024, 1024, 1, 1024 ]
|
|
- Exact: [ 2048, 2048, 1, 2048 ]
|
|
- Exact: [ 4096, 4096, 1, 4096 ]
|
|
- Exact: [ 6144, 6144, 1, 6144]
|
|
- Exact: [ 8192, 8192, 1, 8192 ]
|
|
- Exact: [ 10240, 10240, 1, 10240]
|
|
- Exact: [ 12288, 12288, 1, 12288]
|
|
- Exact: [ 14366, 14366, 1, 14366]
|
|
- Exact: [ 16384, 16384, 1, 16384 ]
|
|
- Exact: [ 18432, 18432, 1, 18432 ]
|
|
- Exact: [ 20480, 20480, 1, 20480 ]
|
|
- Exact: [ 22528, 22528, 1, 22528 ]
|
|
- Exact: [ 24576, 24576, 1, 24576 ]
|
|
- Exact: [ 26624, 26624, 1, 26624 ]
|
|
|
|
- # BenchmarkProblemSizeGroup -
|
|
InitialSolutionParameters:
|
|
BenchmarkCommonParameters:
|
|
- BufferLoad: [True]
|
|
- BufferStore: [True]
|
|
- KernelLanguage: ["Assembly"]
|
|
- EdgeType: ["ShiftPtr"]
|
|
- LoopTail: [True]
|
|
- WorkGroupMapping: [8]
|
|
- PrefetchGlobalRead: [True]
|
|
- VectorWidth: [-1]
|
|
ForkParameters:
|
|
- PrefetchLocalRead: [False]
|
|
- ThreadTile:
|
|
- [ 6, 6 ]
|
|
- WorkGroup:
|
|
- [ 32, 16, 1 ]
|
|
- WorkGroupMapping: [8]
|
|
- GlobalSplitU: [1]
|
|
- DepthU: [ 8 ]
|
|
- VectorWidth: [2]
|
|
- FractionalLoad : [1]
|
|
BenchmarkForkParameters:
|
|
JoinParameters:
|
|
- MacroTile
|
|
- GlobalSplitU
|
|
BenchmarkJoinParameters:
|
|
BenchmarkFinalParameters:
|
|
- ProblemSizes:
|
|
- Exact: [ 16, 16, 1, 16 ]
|
|
- Exact: [ 32, 32, 1, 32 ]
|
|
- Exact: [ 64, 64, 1, 64 ]
|
|
- Exact: [ 128, 128, 1, 128 ]
|
|
- Exact: [ 256, 256, 1, 256 ]
|
|
- Exact: [ 512, 512, 1, 512 ]
|
|
- Exact: [ 1024, 1024, 1, 1024 ]
|
|
- Exact: [ 2048, 2048, 1, 2048 ]
|
|
- Exact: [ 4096, 4096, 1, 4096 ]
|
|
- Exact: [ 6144, 6144, 1, 6144]
|
|
- Exact: [ 8192, 8192, 1, 8192 ]
|
|
- Exact: [ 10240, 10240, 1, 10240]
|
|
- Exact: [ 12288, 12288, 1, 12288]
|
|
- Exact: [ 14366, 14366, 1, 14366]
|
|
- Exact: [ 16384, 16384, 1, 16384 ]
|
|
- Exact: [ 18432, 18432, 1, 18432 ]
|
|
- Exact: [ 20480, 20480, 1, 20480 ]
|
|
- Exact: [ 22528, 22528, 1, 22528 ]
|
|
- Exact: [ 24576, 24576, 1, 24576 ]
|
|
- Exact: [ 26624, 26624, 1, 26624 ]
|
|
|
|
########################################
|