GlobalParameters: {CSVExportWinner: 1, CSVMergeSameProblemID: 1, DataInitTypeAlpha: 1, DataInitTypeBeta: 0, EnqueuesPerSync: 1, KernelTime: true, LegacyComponents: 0, LibraryFormat: yaml, MaxWorkspaceSize: 8388608, MinimumRequiredVersion: 4.14.0, NumBenchmarks: 1, NumElementsToValidate: 0, PrintLevel: 3, ShortNames: 1, SleepPercent: 50, SyncsPerBenchmark: 1, WavefrontWidth: 64, numWarmups: 1} ProblemType: {Batched: true, DataType: s, OperationType: GEMM, TransposeA: false, TransposeB: false, UseBeta: true} BenchmarkCommonParameters: - KernelLanguage: [Assembly] - EdgeType: [ShiftPtr] - ScheduleIterAlg: [0] ForkParameters: - LdcEqualsLdd: [1, 0] - FractionalLoad: [0, 1] - TransposeLDS: [1, 0] - WaveSeparateGlobalReadB: [1, false] - LdsBlockSizePerPad: [0, 128] - WaveSeparateGlobalReadA: [1, false] - StaggerUStride: [256, 128] - PrefetchGlobalRead: [true, 2] - Groups: - - WorkGroup: [8, 8, 1] ThreadTile: [8, 8] - WorkGroup: [16, 16, 1] ThreadTile: [4, 4] - - DepthU: 16 WorkGroupMapping: 1 - DepthU: 16 WorkGroupMapping: 15 - DepthU: 8 WorkGroupMapping: 8