Skip to main content

Ctrl+K

ROCm™ Software Debian Build 7.2.4

Version List

GitHub
Community
Blogs
ROCm Developer Hub
ROCm Toolkits
Systems and Infra Docs
Infinity Hub
Support

ROCm documentation

Composable Kernel 1.2.0 Documentation

Install

Composable Kernel prerequisites
Build and install Composable Kernel
Composable Kernel Docker images

Conceptual

Composable Kernel structure
Composable Kernel mathematical basis

Tutorial

Composable Kernel examples

Reference

Composable Kernel scalar types
Composable Kernel custom types
Composable Kernel vector utilities
Composable Kernel wrapper
CK API reference
CK Tile API reference
Full API class list

About

Contributing to Composable Kernel
License

Composable Kernel: Class Members

Class Members

Class Members#

Composable Kernel: Class Members

Here is a list of all class members with links to the classes they belong to:

- y -

Y0 : ck_tile::tile_distribution_encoding_pattern_2d< BlockSize, YPerTile, XPerTile, VecSize, tile_distribution_pattern::block_raked, NumWaveGroups >, ck_tile::tile_distribution_encoding_pattern_2d< BlockSize, YPerTile, XPerTile, VecSize, tile_distribution_pattern::thread_raked, NumWaveGroups >, ck_tile::tile_distribution_encoding_pattern_2d< BlockSize, YPerTile, XPerTile, VecSize, tile_distribution_pattern::warp_raked, NumWaveGroups >, ck_tile::tile_distribution_encoding_pattern_aq_transposed_c< BlockGemmShape, WarpGemm, BlockSize, YPerTile, XPerTile, VecSize >
Y1 : ck_tile::tile_distribution_encoding_pattern_2d< BlockSize, YPerTile, XPerTile, VecSize, tile_distribution_pattern::block_raked, NumWaveGroups >, ck_tile::tile_distribution_encoding_pattern_2d< BlockSize, YPerTile, XPerTile, VecSize, tile_distribution_pattern::thread_raked, NumWaveGroups >, ck_tile::tile_distribution_encoding_pattern_2d< BlockSize, YPerTile, XPerTile, VecSize, tile_distribution_pattern::warp_raked, NumWaveGroups >, ck_tile::tile_distribution_encoding_pattern_aq_transposed_c< BlockGemmShape, WarpGemm, BlockSize, YPerTile, XPerTile, VecSize >
Y2 : ck_tile::tile_distribution_encoding_pattern_2d< BlockSize, YPerTile, XPerTile, VecSize, tile_distribution_pattern::block_raked, NumWaveGroups >, ck_tile::tile_distribution_encoding_pattern_2d< BlockSize, YPerTile, XPerTile, VecSize, tile_distribution_pattern::thread_raked, NumWaveGroups >, ck_tile::tile_distribution_encoding_pattern_2d< BlockSize, YPerTile, XPerTile, VecSize, tile_distribution_pattern::warp_raked, NumWaveGroups >, ck_tile::tile_distribution_encoding_pattern_aq_transposed_c< BlockGemmShape, WarpGemm, BlockSize, YPerTile, XPerTile, VecSize >
Y_ : ck::tensor_operation::TransformConvBwdDataToGemm_v1< NDimSpatial, ConvBwdDataSpecialization, AK1, BK1, GemmMPerBlock, GemmNPerBlock, GemmKPerBlock, DoPadGemmM, DoPadGemmN, ALayout, BLayout, CLayout, SplitN, ADataType, CDataType, NumGroupsToMerge, IndexType, CTranspose >, ck::tensor_operation::TransformConvFwdToGemm< NDimSpatial, ConvForwardSpecialization, SplitN, ADataType, CDataType, NumGroupsToMerge, IndexType, CTranspose >, ck_tile::TransformConvBwdDataToGemm< NDimSpatial, ConvolutionSpecialization, VectorSizeA, VectorSizeB, VectorSizeC, SplitN, ADataType, CDataType, NumGroupsToMerge, IndexType >, ck_tile::TransformConvBwdWeightToGemm< NDimSpatial, ConvolutionSpecialization, VectorSizeA, VectorSizeB, VectorSizeC, NumGroupsToMerge, SplitN, ADataType, CDataType, IndexType >
y_elementwise_op_ : ck::tensor_operation::device::DeviceBatchNormFwdImpl< XDataType, YDataType, AccDataType, ScaleDataType, BiasDataType, MeanVarDataType, YElementwiseOp, Rank, NumBatchNormReduceDim, UseMultiblockInK, BlockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XSrcYDstVectorDim, XSrcVectorSize, YDstVectorSize, ScaleSrcVectorSize, BiasSrcVectorSize, MeanVarSrcDstVectorSize >::Argument, ck::tensor_operation::device::DeviceElementwiseNormalizationImpl< InDataTypeTuple, GammaDataType, BetaDataType, AccDataType, YDataType, XElementwiseOperation, YElementwiseOperation, Rank, NumReduceDim, BlockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XYSrcVectorDim, XSrcVectorSize, GammaSrcVectorDim, GammaSrcVectorSize, BetaSrcVectorDim, BetaSrcVectorSize, YDstVectorSize >::Argument, ck::tensor_operation::device::DeviceNormalizationFwdImpl< XDataType, GammaDataType, BetaDataType, ComputeDataType, YDataType, SaveMeanInvStdDataType, YElementwiseOperation, Rank, NumReduceDim, BlockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XYSrcVectorDim, XSrcVectorSize, GammaSrcVectorDim, GammaSrcVectorSize, BetaSrcVectorDim, BetaSrcVectorSize, YDstVectorSize, SaveMeanInvStdDstVectorSize, UseWelford >::Argument, ck::tensor_operation::device::DeviceNormalizationFwdSplitKImpl< XDataType, GammaDataType, BetaDataType, ComputeDataType, YDataType, SaveMeanInvStdDataType, YElementwiseOperation, Rank, NumReduceDim, BlockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XYVectorDim, XSrcVectorSize, GammaSrcVectorDim, GammaSrcVectorSize, BetaSrcVectorDim, BetaSrcVectorSize, YDstVectorSize, SaveMeanInvStdDstVectorSize >::Argument
y_grid_desc_m_k_ : ck::tensor_operation::device::DeviceBatchNormFwdImpl< XDataType, YDataType, AccDataType, ScaleDataType, BiasDataType, MeanVarDataType, YElementwiseOp, Rank, NumBatchNormReduceDim, UseMultiblockInK, BlockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XSrcYDstVectorDim, XSrcVectorSize, YDstVectorSize, ScaleSrcVectorSize, BiasSrcVectorSize, MeanVarSrcDstVectorSize >::Argument, ck::tensor_operation::device::DeviceElementwiseNormalizationImpl< InDataTypeTuple, GammaDataType, BetaDataType, AccDataType, YDataType, XElementwiseOperation, YElementwiseOperation, Rank, NumReduceDim, BlockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XYSrcVectorDim, XSrcVectorSize, GammaSrcVectorDim, GammaSrcVectorSize, BetaSrcVectorDim, BetaSrcVectorSize, YDstVectorSize >::Argument, ck::tensor_operation::device::DeviceNormalizationFwdImpl< XDataType, GammaDataType, BetaDataType, ComputeDataType, YDataType, SaveMeanInvStdDataType, YElementwiseOperation, Rank, NumReduceDim, BlockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XYSrcVectorDim, XSrcVectorSize, GammaSrcVectorDim, GammaSrcVectorSize, BetaSrcVectorDim, BetaSrcVectorSize, YDstVectorSize, SaveMeanInvStdDstVectorSize, UseWelford >::Argument, ck::tensor_operation::device::DeviceNormalizationFwdSplitKImpl< XDataType, GammaDataType, BetaDataType, ComputeDataType, YDataType, SaveMeanInvStdDataType, YElementwiseOperation, Rank, NumReduceDim, BlockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XYVectorDim, XSrcVectorSize, GammaSrcVectorDim, GammaSrcVectorSize, BetaSrcVectorDim, BetaSrcVectorSize, YDstVectorSize, SaveMeanInvStdDstVectorSize >::Argument
y_smooth_scale_ptr : ck_tile::FusedMoeGemmHostArgs, ck_tile::FusedMoeGemmKernel< Partitioner_, Pipeline_, Epilogue_ >::FusedMoeGemmKargs
y_stride : ck_tile::Layernorm2dFwd< Pipeline_, Epilogue_ >::Kargs, ck_tile::Layernorm2dFwdHostArgs, ck_tile::MoeSmoothquant< Pipeline_ >::Kargs, ck_tile::MoeSmoothquantHostArgs, ck_tile::Rmsnorm2dFwd< Pipeline_, Epilogue_ >::Kargs, ck_tile::Rmsnorm2dFwdHostArgs, ck_tile::Smoothquant< Pipeline_ >::Kargs, ck_tile::SmoothquantHostArgs
YDataType : ck_tile::ElementWiseKernel< Problem_, Policy_ >, ck_tile::ElementWisePipelineProblem< XDataType_, ComputeDataType_, YDataType_, BlockShape_, ElementWiseOperation_, kPad_ >, ck_tile::FusedMoeGemmKernel< Partitioner_, Pipeline_, Epilogue_ >, ck_tile::FusedMoeGemmPipeline_FlatmmEx< Problem_, Policy_ >, ck_tile::FusedMoeGemmPipeline_FlatmmUk< Problem_, Policy_ >, ck_tile::FusedMoeGemmPipelineProblem< ADataType_, GDataType_, DDataType_, AccDataType_, ODataType_, AScaleDataType_, GScaleDataType_, DScaleDataType_, YSmoothScaleDataType_, TopkWeightDataType_, IndexDataType_, GateActivation_, BlockShape_, Traits_ >, ck_tile::Layernorm2dFwd< Pipeline_, Epilogue_ >, ck_tile::Layernorm2dFwdPipelineOnePass< Problem_, Policy_ >, ck_tile::Layernorm2dFwdPipelineProblem< XDataType_, XBiasDataType_, GammaDataType_, BetaDataType_, ComputeDataType_, YDataType_, MeanDataType_, InvStdDataType_, SmoothScaleDataType_, YScaleDataType_, BlockShape_, Traits_ >, ck_tile::Layernorm2dFwdPipelineTwoPass< Problem_, Policy_ >, ck_tile::Reduce2dProblem< XDataType_, ComputeDataType_, YDataType_, BlockShape_, ReduceOp_, OutputIndex_ >, ck_tile::Reduce< Problem_, Policy_ >, ck_tile::Rmsnorm2dFwd< Pipeline_, Epilogue_ >, ck_tile::Rmsnorm2dFwdPipelineModelSensitiveT5Pass< Problem_, Policy_ >, ck_tile::Rmsnorm2dFwdPipelineOnePass< Problem_, Policy_ >, ck_tile::Rmsnorm2dFwdPipelineProblem< XDataType_, GammaDataType_, ComputeDataType_, YDataType_, InvRmsDataType_, UnquantYDataType_, SmoothScaleDataType_, YScaleDataType_, BlockShape_, Traits_ >, ck_tile::Rmsnorm2dFwdPipelineTwoPass< Problem_, Policy_ >
YDot_ : ck::tensor_operation::TransformConvBwdDataToGemm_v1< NDimSpatial, ConvBwdDataSpecialization, AK1, BK1, GemmMPerBlock, GemmNPerBlock, GemmKPerBlock, DoPadGemmM, DoPadGemmN, ALayout, BLayout, CLayout, SplitN, ADataType, CDataType, NumGroupsToMerge, IndexType, CTranspose >, ck_tile::TransformConvBwdDataToGemm< NDimSpatial, ConvolutionSpecialization, VectorSizeA, VectorSizeB, VectorSizeC, SplitN, ADataType, CDataType, NumGroupsToMerge, IndexType >
YIdx : ck::tensor_operation::device::DeviceColumnToImageImpl< NDimSpatial, ImageLayout, InputDataType, OutputDataType, BlockSize, MPerBlock, KPerBlock, ThreadClusterLengths, ScalarPerVector,, bool, type >
yr_stride : ck_tile::Layernorm2dFwd< Pipeline_, Epilogue_ >::Kargs, ck_tile::Layernorm2dFwdHostArgs, ck_tile::Rmsnorm2dFwd< Pipeline_, Epilogue_ >::Kargs, ck_tile::Rmsnorm2dFwdHostArgs
YResidualDataType : ck_tile::Layernorm2dFwd< Pipeline_, Epilogue_ >, ck_tile::Layernorm2dFwdPipelineOnePass< Problem_, Policy_ >, ck_tile::Layernorm2dFwdPipelineTwoPass< Problem_, Policy_ >, ck_tile::Rmsnorm2dFwd< Pipeline_, Epilogue_ >, ck_tile::Rmsnorm2dFwdPipelineModelSensitiveT5Pass< Problem_, Policy_ >, ck_tile::Rmsnorm2dFwdPipelineOnePass< Problem_, Policy_ >, ck_tile::Rmsnorm2dFwdPipelineTwoPass< Problem_, Policy_ >
Ys2DDescriptor : ck_tile::tile_distribution< PsYs2XsAdaptor_, Ys2DDescriptor_, StaticTileDistributionEncoding_, TileDistributionDetail_ >
Ys2RHsMajor : ck_tile::tile_distribution_encoding< RsLengths_, HsLengthss_, Ps2RHssMajor_, Ps2RHssMinor_, Ys2RHsMajor_, Ys2RHsMinor_ >
Ys2RHsMinor : ck_tile::tile_distribution_encoding< RsLengths_, HsLengthss_, Ps2RHssMajor_, Ps2RHssMinor_, Ys2RHsMajor_, Ys2RHsMinor_ >
ys_lengths_ : ck_tile::tile_distribution_encoding< RsLengths_, HsLengthss_, Ps2RHssMajor_, Ps2RHssMinor_, Ys2RHsMajor_, Ys2RHsMinor_ >::detail
ys_mapping_valid : ck_tile::DefaultTranspose< DataType >::ValidationTraitsImpl< InDstrEncode, ReverseDirection, LaneGroupSize >
ys_to_d_ : ck_tile::tile_distribution< PsYs2XsAdaptor_, Ys2DDescriptor_, StaticTileDistributionEncoding_, TileDistributionDetail_ >
ys_to_rhs_major_ : ck_tile::tile_distribution_encoding< RsLengths_, HsLengthss_, Ps2RHssMajor_, Ps2RHssMinor_, Ys2RHsMajor_, Ys2RHsMinor_ >
ys_to_rhs_minor_ : ck_tile::tile_distribution_encoding< RsLengths_, HsLengthss_, Ps2RHssMajor_, Ps2RHssMinor_, Ys2RHsMajor_, Ys2RHsMinor_ >
ys_to_span_major_ : ck_tile::tile_distribution_encoding< RsLengths_, HsLengthss_, Ps2RHssMajor_, Ps2RHssMinor_, Ys2RHsMajor_, Ys2RHsMinor_ >::detail
ys_to_span_minor_ : ck_tile::tile_distribution_encoding< RsLengths_, HsLengthss_, Ps2RHssMajor_, Ps2RHssMinor_, Ys2RHsMajor_, Ys2RHsMinor_ >::detail
YScaleDataType : ck_tile::AddRmsnorm2dRdquantFwd< Pipeline_ >, ck_tile::AddRmsnorm2dRdquantFwdPipelineOnePass< Problem_, Policy_ >, ck_tile::AddRmsnorm2dRdquantFwdPipelineProblem< ADataType_, BDataType_, GammaDataType_, ComputeDataType_, XDataType_, YScaleDataType_, QYDataType_, BlockShape_, kPadN_, kSaveX_, kThreePass_ >, ck_tile::AddRmsnorm2dRdquantFwdPipelineThreePass< Problem_, Policy_ >, ck_tile::Default2DAndDynamicQuantEpilogueProblem< AccDataType_, SmoothScaleDataType_, YScaleDataType_, ODataType_, UnquantYDataType_, BlockShape_, Traits_ >, ck_tile::DynamicQuantEpilogue< Problem_, Policy_ >, ck_tile::DynamicQuantEpilogueProblem< AccDataType_, SmoothScaleDataType_, YScaleDataType_, ODataType_, BlockShape_, Traits_ >, ck_tile::Layernorm2dFwd< Pipeline_, Epilogue_ >, ck_tile::Layernorm2dFwdPipelineProblem< XDataType_, XBiasDataType_, GammaDataType_, BetaDataType_, ComputeDataType_, YDataType_, MeanDataType_, InvStdDataType_, SmoothScaleDataType_, YScaleDataType_, BlockShape_, Traits_ >, ck_tile::MoeSmoothquant< Pipeline_ >, ck_tile::Rmsnorm2dFwd< Pipeline_, Epilogue_ >, ck_tile::Rmsnorm2dFwdPipelineProblem< XDataType_, GammaDataType_, ComputeDataType_, YDataType_, InvRmsDataType_, UnquantYDataType_, SmoothScaleDataType_, YScaleDataType_, BlockShape_, Traits_ >, ck_tile::Smoothquant< Pipeline_ >, ck_tile::SmoothquantPipelineOnePass< Problem_, Policy_ >, ck_tile::SmoothquantPipelineProblem< XDataType_, SmoothScaleDataType_, ComputeDataType_, YScaleDataType_, QYDataType_, BlockShape_, kPadN_, kTwoPass_ >, ck_tile::SmoothquantPipelineTwoPass< Problem_, Policy_ >
YSmoothScaleDataType : ck_tile::FusedMoeGemmKernel< Partitioner_, Pipeline_, Epilogue_ >, ck_tile::FusedMoeGemmPipeline_FlatmmEx< Problem_, Policy_ >, ck_tile::FusedMoeGemmPipeline_FlatmmUk< Problem_, Policy_ >, ck_tile::FusedMoeGemmPipelineProblem< ADataType_, GDataType_, DDataType_, AccDataType_, ODataType_, AScaleDataType_, GScaleDataType_, DScaleDataType_, YSmoothScaleDataType_, TopkWeightDataType_, IndexDataType_, GateActivation_, BlockShape_, Traits_ >
yStrides_ : ck::tensor_operation::device::DeviceBatchNormFwdImpl< XDataType, YDataType, AccDataType, ScaleDataType, BiasDataType, MeanVarDataType, YElementwiseOp, Rank, NumBatchNormReduceDim, UseMultiblockInK, BlockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XSrcYDstVectorDim, XSrcVectorSize, YDstVectorSize, ScaleSrcVectorSize, BiasSrcVectorSize, MeanVarSrcDstVectorSize >::Argument, ck::tensor_operation::device::DeviceElementwiseNormalizationImpl< InDataTypeTuple, GammaDataType, BetaDataType, AccDataType, YDataType, XElementwiseOperation, YElementwiseOperation, Rank, NumReduceDim, BlockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XYSrcVectorDim, XSrcVectorSize, GammaSrcVectorDim, GammaSrcVectorSize, BetaSrcVectorDim, BetaSrcVectorSize, YDstVectorSize >::Argument, ck::tensor_operation::device::DeviceNormalizationFwdImpl< XDataType, GammaDataType, BetaDataType, ComputeDataType, YDataType, SaveMeanInvStdDataType, YElementwiseOperation, Rank, NumReduceDim, BlockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XYSrcVectorDim, XSrcVectorSize, GammaSrcVectorDim, GammaSrcVectorSize, BetaSrcVectorDim, BetaSrcVectorSize, YDstVectorSize, SaveMeanInvStdDstVectorSize, UseWelford >::Argument, ck::tensor_operation::device::DeviceNormalizationFwdSplitKImpl< XDataType, GammaDataType, BetaDataType, ComputeDataType, YDataType, SaveMeanInvStdDataType, YElementwiseOperation, Rank, NumReduceDim, BlockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XYVectorDim, XSrcVectorSize, GammaSrcVectorDim, GammaSrcVectorSize, BetaSrcVectorDim, BetaSrcVectorSize, YDstVectorSize, SaveMeanInvStdDstVectorSize >::Argument
YThreadBufferNumber : ck::GridwiseElementwiseLayernormWelfordVariance_mk_to_mk< InDataTypePointerTuple, XDataType, GammaDataType, BetaDataType, YDataType, AccDataType, XElementwiseOperation, YElementwiseOperation, InGrid2dDescTuple, GridDesc_M_K, BlockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XSrcVectorDim, XSrcVectorSize, GammaSrcVectorDim, GammaSrcVectorSize, BetaSrcVectorDim, BetaSrcVectorSize, YDstVectorDim, YDstVectorSize, SweepOnce >
YTilde_ : ck::tensor_operation::TransformConvBwdDataToGemm_v1< NDimSpatial, ConvBwdDataSpecialization, AK1, BK1, GemmMPerBlock, GemmNPerBlock, GemmKPerBlock, DoPadGemmM, DoPadGemmN, ALayout, BLayout, CLayout, SplitN, ADataType, CDataType, NumGroupsToMerge, IndexType, CTranspose >, ck_tile::TransformConvBwdDataToGemm< NDimSpatial, ConvolutionSpecialization, VectorSizeA, VectorSizeB, VectorSizeC, SplitN, ADataType, CDataType, NumGroupsToMerge, IndexType >

Terms and Conditions
ROCm Licenses and Disclaimers
Privacy
Trademarks
Supply Chain Transparency
Fair and Open Competition
UK Tax Strategy
Cookie Policy
Cookie Settings

© 2026 Advanced Micro Devices, Inc