- e -

• Element : cutlass::Array< T, N, false > , cutlass::Array< T, N, true > , cutlass::ConstSubbyteReference< Element_, Storage_ > , cutlass::epilogue::thread::ReductionOpPlus< Element_, Count > , cutlass::epilogue::threadblock::DefaultInterleavedThreadMapTensorOp< ThreadblockShape_, WarpShape_, PartitionsK, Element_, ElementsPerAccess, InterleavedK > , cutlass::epilogue::threadblock::DefaultThreadMapSimt< ThreadblockShape_, WarpShape_, MmaSimtPolicy_, PartitionsK, Element_, ElementsPerAccess > , cutlass::epilogue::threadblock::DefaultThreadMapTensorOp< ThreadblockShape_, WarpShape_, PartitionsK, Element_, ElementsPerAccess > , cutlass::epilogue::threadblock::DefaultThreadMapWmmaTensorOp< ThreadblockShape_, WarpShape_, InstructionShape_, PartitionsK, Element_, ElementsPerAccess > , cutlass::epilogue::threadblock::DirectEpilogueTensorOp< Shape_, Operator_, PartitionsK, Element_, OutputOp_, ConvertOp_ > , cutlass::epilogue::threadblock::EpilogueBase< Shape_, WarpMmaOperator_, PartitionsK, AccumulatorFragmentIterator_, WarpTileIterator_, Padding_ >::SharedStorage , cutlass::epilogue::threadblock::InterleavedPredicatedTileIterator< ThreadMap_, Element_, InterleavedK > , cutlass::epilogue::threadblock::PredicatedTileIterator< ThreadMap_, Element_ > , cutlass::epilogue::threadblock::SharedLoadIterator< ThreadMap_, Element_, MaxAlignment > , cutlass::epilogue::warp::TileIteratorSimt< WarpShape_, Operator_, Element_, layout::RowMajor, MmaSimtPolicy_ > , cutlass::epilogue::warp::TileIteratorTensorOp< WarpShape_, OperatorShape_, Element_, layout::RowMajor > , cutlass::epilogue::warp::TileIteratorVoltaTensorOp< WarpShape_, gemm::GemmShape< 32, 32, 4 >, float, layout::RowMajor > , cutlass::epilogue::warp::TileIteratorVoltaTensorOp< WarpShape_, gemm::GemmShape< 32, 32, 4 >, half_t, layout::RowMajor > , cutlass::epilogue::warp::TileIteratorWmmaTensorOp< WarpShape_, OperatorShape_, OperatorFragment_, layout::RowMajor > , cutlass::gemm::warp::MmaSimtTileIterator< Shape_, Operand::kA, Element_, layout::ColumnMajor, Policy_, PartitionsK, PartitionGroupSize > , cutlass::gemm::warp::MmaSimtTileIterator< Shape_, Operand::kA, Element_, layout::ColumnMajorInterleaved< 4 >, Policy_, PartitionsK, PartitionGroupSize > , cutlass::gemm::warp::MmaSimtTileIterator< Shape_, Operand::kB, Element_, layout::RowMajor, Policy_, PartitionsK, PartitionGroupSize > , cutlass::gemm::warp::MmaSimtTileIterator< Shape_, Operand::kB, Element_, layout::RowMajorInterleaved< 4 >, Policy_, PartitionsK, PartitionGroupSize > , cutlass::gemm::warp::MmaSimtTileIterator< Shape_, Operand::kC, Element_, layout::ColumnMajor, Policy_ > , cutlass::gemm::warp::MmaSimtTileIterator< Shape_, Operand::kC, Element_, layout::RowMajor, Policy_ > , cutlass::gemm::warp::MmaTensorOpAccumulatorTileIterator< Shape_, Element_, cutlass::layout::ColumnMajor, InstructionShape_, OpDelta_ > , cutlass::gemm::warp::MmaTensorOpAccumulatorTileIterator< Shape_, Element_, cutlass::layout::ColumnMajorInterleaved< InterleavedN >, InstructionShape_, OpDelta_ > , cutlass::gemm::warp::MmaTensorOpAccumulatorTileIterator< Shape_, Element_, cutlass::layout::RowMajor, InstructionShape_, OpDelta_ > , cutlass::gemm::warp::MmaTensorOpMultiplicandTileIterator< Shape_, Operand_, Element_, cutlass::layout::ColumnMajorTensorOpMultiplicandCongruous< sizeof_bits< Element_ >::value, int(128/sizeof(Element_))>, InstructionShape_, OpDelta_, 32, PartitionsK_ > , cutlass::gemm::warp::MmaTensorOpMultiplicandTileIterator< Shape_, Operand_, Element_, cutlass::layout::ColumnMajorTensorOpMultiplicandCrosswise< sizeof_bits< Element_ >::value, Crosswise >, InstructionShape_, OpDelta_, 32, PartitionsK_ > , cutlass::gemm::warp::MmaTensorOpMultiplicandTileIterator< Shape_, Operand_, Element_, cutlass::layout::RowMajorTensorOpMultiplicandCongruous< sizeof_bits< Element_ >::value, int(128/sizeof(Element_))>, InstructionShape_, OpDelta_, 32, PartitionsK_ > , cutlass::gemm::warp::MmaTensorOpMultiplicandTileIterator< Shape_, Operand_, Element_, cutlass::layout::RowMajorTensorOpMultiplicandCrosswise< sizeof_bits< Element_ >::value, Crosswise >, InstructionShape_, OpDelta_, 32, PartitionsK_ > , cutlass::gemm::warp::MmaTensorOpMultiplicandTileIterator< Shape_, Operand_, Element_, cutlass::layout::TensorOpMultiplicandCongruous< sizeof_bits< Element_ >::value, 64 >, InstructionShape_, OpDelta_, 32, PartitionsK_ > , cutlass::gemm::warp::MmaTensorOpMultiplicandTileIterator< Shape_, Operand_, Element_, cutlass::layout::TensorOpMultiplicandCrosswise< sizeof_bits< Element_ >::value, Crosswise >, InstructionShape_, OpDelta_, 32, PartitionsK_ > , cutlass::gemm::warp::MmaVoltaTensorOpAccumulatorTileIterator< Shape_, Element_, Layout_, InstructionShape_, OpDelta_ > , cutlass::gemm::warp::MmaVoltaTensorOpMultiplicandTileIterator< Shape_, Operand::kA, Element_, cutlass::layout::ColumnMajorVoltaTensorOpMultiplicandCongruous< sizeof_bits< Element_ >::value >, InstructionShape_, OpDelta_, 32 > , cutlass::gemm::warp::MmaVoltaTensorOpMultiplicandTileIterator< Shape_, Operand::kA, Element_, cutlass::layout::VoltaTensorOpMultiplicandCongruous< sizeof_bits< Element_ >::value >, InstructionShape_, OpDelta_, 32 > , cutlass::gemm::warp::MmaVoltaTensorOpMultiplicandTileIterator< Shape_, Operand::kB, Element_, cutlass::layout::RowMajorVoltaTensorOpMultiplicandBCongruous< sizeof_bits< Element_ >::value >, InstructionShape_, OpDelta_, 32 > , cutlass::gemm::warp::MmaVoltaTensorOpMultiplicandTileIterator< Shape_, Operand::kB, Element_, cutlass::layout::VoltaTensorOpMultiplicandBCongruous< sizeof_bits< Element_ >::value >, InstructionShape_, OpDelta_, 32 > , cutlass::gemm::warp::MmaVoltaTensorOpMultiplicandTileIterator< Shape_, Operand_, Element_, cutlass::layout::ColumnMajorVoltaTensorOpMultiplicandCrosswise< sizeof_bits< Element_ >::value, KBlock >, InstructionShape_, OpDelta_, 32 > , cutlass::gemm::warp::MmaVoltaTensorOpMultiplicandTileIterator< Shape_, Operand_, Element_, cutlass::layout::RowMajorVoltaTensorOpMultiplicandCrosswise< sizeof_bits< Element_ >::value, KBlock >, InstructionShape_, OpDelta_, 32 > , cutlass::gemm::warp::MmaVoltaTensorOpMultiplicandTileIterator< Shape_, Operand_, Element_, cutlass::layout::VoltaTensorOpMultiplicandCrosswise< sizeof_bits< Element_ >::value, KBlock >, InstructionShape_, OpDelta_, 32 > , cutlass::HostTensor< Element_, Layout_ > , cutlass::reduction::thread::ReduceAdd< ElementAccumulator_, Element_, Count > , cutlass::SubbyteReference< Element_, Storage_ > , cutlass::TensorRef< Element_, Layout_ > , cutlass::TensorView< Element_, Layout_ > , cutlass::thread::Matrix< Element, Rows, Columns, Layout > , cutlass::transform::thread::Transpose< ElementCount_, layout::PitchLinearShape< 4, 4 >, int8_t > , cutlass::transform::threadblock::PredicatedTileAccessIterator2dThreadTile< Shape_, Element_, layout::ColumnMajor, AdvanceRank, ThreadMap_, AccessType_ > , cutlass::transform::threadblock::PredicatedTileAccessIterator2dThreadTile< Shape_, Element_, layout::PitchLinear, AdvanceRank, ThreadMap_, AccessType_ > , cutlass::transform::threadblock::PredicatedTileAccessIterator2dThreadTile< Shape_, Element_, layout::RowMajor, AdvanceRank, ThreadMap_, AccessType_ > , cutlass::transform::threadblock::PredicatedTileAccessIterator< Shape_, Element_, layout::ColumnMajor, AdvanceRank, ThreadMap_, AccessType_ > , cutlass::transform::threadblock::PredicatedTileAccessIterator< Shape_, Element_, layout::ColumnMajorInterleaved< InterleavedK >, AdvanceRank, ThreadMap_, AccessType_ > , cutlass::transform::threadblock::PredicatedTileAccessIterator< Shape_, Element_, layout::PitchLinear, AdvanceRank, ThreadMap_, AccessType_ > , cutlass::transform::threadblock::PredicatedTileAccessIterator< Shape_, Element_, layout::RowMajor, AdvanceRank, ThreadMap_, AccessType_ > , cutlass::transform::threadblock::PredicatedTileAccessIterator< Shape_, Element_, layout::RowMajorInterleaved< InterleavedK >, AdvanceRank, ThreadMap_, AccessType_ > , cutlass::transform::threadblock::PredicatedTileIterator2dThreadTile< Shape_, Element_, layout::ColumnMajor, AdvanceRank, ThreadMap_, Transpose_ > , cutlass::transform::threadblock::PredicatedTileIterator2dThreadTile< Shape_, Element_, layout::PitchLinear, AdvanceRank, ThreadMap_, Transpose_ > , cutlass::transform::threadblock::PredicatedTileIterator2dThreadTile< Shape_, Element_, layout::RowMajor, AdvanceRank, ThreadMap_, Transpose_ > , cutlass::transform::threadblock::PredicatedTileIterator< Shape_, Element_, layout::ColumnMajor, AdvanceRank, ThreadMap_, AccessSize > , cutlass::transform::threadblock::PredicatedTileIterator< Shape_, Element_, layout::ColumnMajorInterleaved< InterleavedK >, AdvanceRank, ThreadMap_, AccessSize > , cutlass::transform::threadblock::PredicatedTileIterator< Shape_, Element_, layout::PitchLinear, AdvanceRank, ThreadMap_, AccessSize > , cutlass::transform::threadblock::PredicatedTileIterator< Shape_, Element_, layout::RowMajor, AdvanceRank, ThreadMap_, AccessSize > , cutlass::transform::threadblock::PredicatedTileIterator< Shape_, Element_, layout::RowMajorInterleaved< InterleavedK >, AdvanceRank, ThreadMap_, AccessSize > , cutlass::transform::threadblock::RegularTileAccessIterator< Shape_, Element_, layout::ColumnMajor, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileAccessIterator< Shape_, Element_, layout::ColumnMajorTensorOpMultiplicandCongruous< sizeof_bits< Element_ >::value, int(128/sizeof(Element_))>, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileAccessIterator< Shape_, Element_, layout::ColumnMajorTensorOpMultiplicandCrosswise< sizeof_bits< Element_ >::value, Crosswise >, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileAccessIterator< Shape_, Element_, layout::PitchLinear, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileAccessIterator< Shape_, Element_, layout::RowMajor, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileAccessIterator< Shape_, Element_, layout::RowMajorTensorOpMultiplicandCongruous< sizeof_bits< Element_ >::value, int(128/sizeof(Element_))>, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileAccessIterator< Shape_, Element_, layout::RowMajorTensorOpMultiplicandCrosswise< sizeof_bits< Element_ >::value, Crosswise >, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileAccessIterator< Shape_, Element_, layout::TensorOpMultiplicandCongruous< sizeof_bits< Element_ >::value, int(128/sizeof(Element_))>, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileAccessIterator< Shape_, Element_, layout::TensorOpMultiplicandCrosswise< sizeof_bits< Element_ >::value, Crosswise >, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator2dThreadTile< Shape_, Element_, layout::ColumnMajorInterleaved< 4 >, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator2dThreadTile< Shape_, Element_, layout::PitchLinear, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator2dThreadTile< Shape_, Element_, layout::RowMajorInterleaved< 4 >, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::ColumnMajor, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::ColumnMajorTensorOpMultiplicandCongruous< sizeof_bits< Element_ >::value, int(128/sizeof(Element_))>, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::ColumnMajorTensorOpMultiplicandCrosswise< sizeof_bits< Element_ >::value, Crosswise >, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::ColumnMajorVoltaTensorOpMultiplicandBCongruous< sizeof_bits< Element_ >::value >, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::ColumnMajorVoltaTensorOpMultiplicandCongruous< sizeof_bits< Element_ >::value >, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::ColumnMajorVoltaTensorOpMultiplicandCrosswise< sizeof_bits< Element_ >::value, Shape_::kRow >, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::PitchLinear, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::RowMajor, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::RowMajorTensorOpMultiplicandCongruous< sizeof_bits< Element_ >::value, int(128/sizeof(Element_))>, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::RowMajorTensorOpMultiplicandCrosswise< sizeof_bits< Element_ >::value, Crosswise >, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::RowMajorVoltaTensorOpMultiplicandBCongruous< sizeof_bits< Element_ >::value >, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::RowMajorVoltaTensorOpMultiplicandCongruous< sizeof_bits< Element_ >::value >, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::RowMajorVoltaTensorOpMultiplicandCrosswise< sizeof_bits< Element_ >::value, Shape_::kColumn >, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::TensorOpMultiplicandCongruous< sizeof_bits< Element_ >::value, int(128/sizeof(Element_))>, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::TensorOpMultiplicandCrosswise< sizeof_bits< Element_ >::value, Crosswise >, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::VoltaTensorOpMultiplicandBCongruous< sizeof_bits< Element_ >::value >, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::VoltaTensorOpMultiplicandCongruous< sizeof_bits< Element_ >::value >, AdvanceRank, ThreadMap_, Alignment > , cutlass::transform::threadblock::RegularTileIterator< Shape_, Element_, layout::VoltaTensorOpMultiplicandCrosswise< sizeof_bits< Element_ >::value, Shape_::kContiguous >, AdvanceRank, ThreadMap_, Alignment >
• element_type : cutlass::platform::unique_ptr< T, Deleter >
• ElementA : cutlass::arch::Mma< gemm::GemmShape< 16, 8, 8 >, 32, half_t, layout::RowMajor, half_t, layout::ColumnMajor, float, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 16, 8, 8 >, 32, half_t, layout::RowMajor, half_t, layout::ColumnMajor, half_t, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 128 >, 32, uint1b_t, layout::RowMajor, uint1b_t, layout::ColumnMajor, int, layout::RowMajor, OpXorPopc > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, int8_t, layout::RowMajor, int8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, int8_t, layout::RowMajor, int8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, int8_t, layout::RowMajor, uint8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, int8_t, layout::RowMajor, uint8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, uint8_t, layout::RowMajor, int8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, uint8_t, layout::RowMajor, int8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, uint8_t, layout::RowMajor, uint8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, uint8_t, layout::RowMajor, uint8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, int4b_t, layout::RowMajor, int4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, int4b_t, layout::RowMajor, int4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, int4b_t, layout::RowMajor, uint4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, int4b_t, layout::RowMajor, uint4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, uint4b_t, layout::RowMajor, int4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, uint4b_t, layout::RowMajor, int4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, uint4b_t, layout::RowMajor, uint4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, uint4b_t, layout::RowMajor, uint4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::ColumnMajor, half_t, layout::ColumnMajor, float, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::ColumnMajor, half_t, layout::ColumnMajor, half_t, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::ColumnMajor, half_t, layout::RowMajor, float, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::ColumnMajor, half_t, layout::RowMajor, half_t, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::RowMajor, half_t, layout::ColumnMajor, float, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::RowMajor, half_t, layout::ColumnMajor, half_t, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::RowMajor, half_t, layout::RowMajor, float, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::RowMajor, half_t, layout::RowMajor, half_t, layout::RowMajor, OpMultiplyAdd > , cutlass::gemm::device::Gemm< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, SplitKSerial, Operator_, IsBetaZero > , cutlass::gemm::device::Gemm< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, SplitKSerial, Operator_, IsBetaZero > , cutlass::gemm::device::GemmBatched< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, Operator_ > , cutlass::gemm::device::GemmBatched< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, Operator_ > , cutlass::gemm::device::GemmComplex< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, TransformA, TransformB, SplitKSerial > , cutlass::gemm::device::GemmComplex< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, TransformA, TransformB, SplitKSerial > , cutlass::gemm::device::GemmSplitKParallel< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ConvertScaledOp_, ReductionOp_, ThreadblockSwizzle_, Stages, kAlignmentA, kAlignmentB, Operator_ > , cutlass::gemm::device::GemmSplitKParallel< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ConvertScaledOp_, ReductionOp_, ThreadblockSwizzle_, Stages, kAlignmentA, kAlignmentB, Operator_ > , cutlass::gemm::kernel::DefaultGemm< int8_t, LayoutA, kAlignmentA, int8_t, LayoutB, kAlignmentB, ElementC, LayoutC, ElementAccumulator, arch::OpClassSimt, ArchTag, ThreadblockShape, WarpShape, GemmShape< 1, 1, 4 >, EpilogueOutputOp, ThreadblockSwizzle, 2, SplitKSerial, Operator, false > , cutlass::gemm::kernel::DefaultGemv< ThreadBlockShape_, ThreadShape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementCD_, LayoutCD_, ElementAccumulator_ > , cutlass::gemm::thread::Mma< Shape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, arch::OpMultiplyAdd, bool > , cutlass::gemm::thread::Mma< Shape_, half_t, LayoutA, half_t, LayoutB, half_t, LayoutC, arch::OpMultiplyAdd > , cutlass::gemm::thread::Mma< Shape_, half_t, LayoutA_, half_t, LayoutB_, half_t, layout::RowMajor, arch::OpMultiplyAdd, typename platform::enable_if< detail::EnableMma_Crow_SM60< LayoutA_, LayoutB_ >::value >::type > , cutlass::gemm::thread::Mma< Shape_, int8_t, layout::ColumnMajor, int8_t, layout::RowMajor, int32_t, LayoutC_, arch::OpMultiplyAdd, int8_t > , cutlass::gemm::thread::Mma< Shape_, int8_t, layout::RowMajor, int8_t, layout::ColumnMajor, int32_t, LayoutC_, arch::OpMultiplyAdd, bool > , cutlass::gemm::thread::MmaGeneric< Shape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, Operator_ > , cutlass::gemm::threadblock::DefaultGemvCore< Shape_, ThreadShape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_ > , cutlass::gemm::threadblock::DefaultMma< int8_t, LayoutA, kAlignmentA, int8_t, LayoutB, kAlignmentB, ElementAccumulator, layout::RowMajor, arch::OpClassSimt, ArchTag, ThreadblockShape, WarpShape, GemmShape< 1, 1, 4 >, 2, Operator, false > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 1 >, ElementA_, layout::ColumnMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 1 >, ElementA_, layout::ColumnMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 1 >, ElementA_, layout::ColumnMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_, > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 1 >, ElementA_, layout::RowMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 1 >, ElementA_, layout::RowMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 4 >, int8_t, layout::ColumnMajor, int8_t, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 4 >, int8_t, layout::ColumnMajor, int8_t, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 4 >, int8_t, layout::RowMajor, int8_t, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 4 >, int8_t, layout::RowMajor, int8_t, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 8, 8, 4 >, ElementA_, layout::ColumnMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 8, 8, 4 >, ElementA_, layout::ColumnMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 8, 8, 4 >, ElementA_, layout::RowMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 8, 8, 4 >, ElementA_, layout::RowMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, InstructionShape_, ElementA_, layout::ColumnMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, InstructionShape_, ElementA_, layout::ColumnMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, InstructionShape_, ElementA_, layout::ColumnMajorInterleaved< InterleavedK >, ElementB_, layout::RowMajorInterleaved< InterleavedK >, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_, AccumulatorsInRowMajor > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, InstructionShape_, ElementA_, layout::RowMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, InstructionShape_, ElementA_, layout::RowMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::warp::MmaComplexTensorOp< Shape_, complex< RealElementA >, LayoutA_, complex< RealElementB >, LayoutB_, complex< RealElementC >, LayoutC_, Policy_, TransformA, TransformB, Enable > , cutlass::gemm::warp::MmaSimt< Shape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, Policy_, PartitionsK, Enable > , cutlass::gemm::warp::MmaTensorOp< Shape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, Policy_, PartitionsK_, AccumulatorsInRowMajor, PartitionsN_, Enable > , cutlass::gemm::warp::MmaVoltaTensorOp< Shape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, Policy_, Enable > , cutlass::reference::device::thread::Gemm< TensorRefA, TensorRefB, TensorRefC, ScalarType, AccumulatorType, OutputTile, InnerProductOp, ConvertOp >
• ElementAccumulator : cutlass::epilogue::thread::Convert< ElementOutput_, Count, ElementAccumulator_, Round > , cutlass::epilogue::thread::LinearCombination< ElementOutput_, Count, ElementAccumulator_, ElementCompute_, Round > , cutlass::epilogue::thread::LinearCombinationClamp< ElementOutput_, Count, ElementAccumulator_, ElementCompute_, Round > , cutlass::epilogue::thread::LinearCombinationRelu< ElementOutput_, Count, ElementAccumulator_, ElementCompute_, Round > , cutlass::epilogue::thread::LinearCombinationRelu< ElementOutput_, Count, int, float, Round > , cutlass::epilogue::threadblock::DefaultEpilogueComplexTensorOp< Shape_, WarpMmaTensorOp_, PartitionsK, OutputOp_, ElementsPerAccess > , cutlass::epilogue::threadblock::DefaultEpilogueSimt< Shape_, WarpMmaSimt_, OutputOp_, ElementsPerAccess > , cutlass::epilogue::threadblock::DefaultEpilogueTensorOp< Shape_, WarpMmaTensorOp_, PartitionsK, OutputOp_, ElementsPerAccess > , cutlass::epilogue::threadblock::DefaultEpilogueVoltaTensorOp< Shape_, WarpMmaTensorOp_, PartitionsK, OutputOp_, ElementsPerAccess > , cutlass::epilogue::threadblock::DefaultEpilogueWmmaTensorOp< Shape_, WarpMmaTensorOp_, PartitionsK, OutputOp_, ElementsPerAccess > , cutlass::epilogue::threadblock::DefaultInterleavedEpilogueTensorOp< Shape_, WarpMmaTensorOp_, PartitionsK, OutputOp_, ElementsPerAccess, InterleavedK, IsBetaZero, isSplitK > , cutlass::epilogue::threadblock::DefaultThreadMapVoltaTensorOp< ThreadblockShape_, WarpShape_, PartitionsK, ElementOutput_, ElementsPerAccess, float > , cutlass::epilogue::threadblock::DefaultThreadMapVoltaTensorOp< ThreadblockShape_, WarpShape_, PartitionsK, ElementOutput_, ElementsPerAccess, half_t > , cutlass::epilogue::threadblock::Epilogue< Shape_, WarpMmaOperator_, PartitionsK, OutputTileIterator_, AccumulatorFragmentIterator_, WarpTileIterator_, SharedLoadIterator_, OutputOp_, Padding_ > , cutlass::epilogue::threadblock::EpilogueBase< Shape_, WarpMmaOperator_, PartitionsK, AccumulatorFragmentIterator_, WarpTileIterator_, Padding_ > , cutlass::epilogue::threadblock::InterleavedEpilogue< Shape_, WarpMmaOperator_, PartitionsK, OutputTileIterator_, AccumulatorFragmentIterator_, OutputOp_, InterleavedK, IsBetaZero > , cutlass::gemm::device::Gemm< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, SplitKSerial, Operator_, IsBetaZero > , cutlass::gemm::device::Gemm< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, SplitKSerial, Operator_, IsBetaZero > , cutlass::gemm::device::GemmBatched< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, Operator_ > , cutlass::gemm::device::GemmBatched< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, Operator_ > , cutlass::gemm::device::GemmComplex< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, TransformA, TransformB, SplitKSerial > , cutlass::gemm::device::GemmComplex< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, TransformA, TransformB, SplitKSerial > , cutlass::gemm::device::GemmSplitKParallel< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ConvertScaledOp_, ReductionOp_, ThreadblockSwizzle_, Stages, kAlignmentA, kAlignmentB, Operator_ > , cutlass::gemm::device::GemmSplitKParallel< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ConvertScaledOp_, ReductionOp_, ThreadblockSwizzle_, Stages, kAlignmentA, kAlignmentB, Operator_ > , cutlass::gemm::kernel::DefaultGemm< ElementA, layout::ColumnMajorInterleaved< InterleavedK >, kAlignmentA, ElementB, layout::RowMajorInterleaved< InterleavedK >, kAlignmentB, ElementC, layout::ColumnMajorInterleaved< InterleavedK >, int32_t, arch::OpClassTensorOp, arch::Sm75, ThreadblockShape, WarpShape, InstructionShape, EpilogueOutputOp, ThreadblockSwizzle, 2, SplitKSerial, Operator, IsBetaZero > , cutlass::gemm::kernel::DefaultGemv< ThreadBlockShape_, ThreadShape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementCD_, LayoutCD_, ElementAccumulator_ > , cutlass::reduction::kernel::ReduceSplitK< Shape_, OutputOp_, ReductionOp_, PartitionsPerStage > , cutlass::reduction::thread::ReduceAdd< ElementAccumulator_, Element_, Count >
• ElementB : cutlass::arch::Mma< gemm::GemmShape< 16, 8, 8 >, 32, half_t, layout::RowMajor, half_t, layout::ColumnMajor, float, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 16, 8, 8 >, 32, half_t, layout::RowMajor, half_t, layout::ColumnMajor, half_t, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 128 >, 32, uint1b_t, layout::RowMajor, uint1b_t, layout::ColumnMajor, int, layout::RowMajor, OpXorPopc > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, int8_t, layout::RowMajor, int8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, int8_t, layout::RowMajor, int8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, int8_t, layout::RowMajor, uint8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, int8_t, layout::RowMajor, uint8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, uint8_t, layout::RowMajor, int8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, uint8_t, layout::RowMajor, int8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, uint8_t, layout::RowMajor, uint8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, uint8_t, layout::RowMajor, uint8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, int4b_t, layout::RowMajor, int4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, int4b_t, layout::RowMajor, int4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, int4b_t, layout::RowMajor, uint4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, int4b_t, layout::RowMajor, uint4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, uint4b_t, layout::RowMajor, int4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, uint4b_t, layout::RowMajor, int4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, uint4b_t, layout::RowMajor, uint4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, uint4b_t, layout::RowMajor, uint4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::ColumnMajor, half_t, layout::ColumnMajor, float, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::ColumnMajor, half_t, layout::ColumnMajor, half_t, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::ColumnMajor, half_t, layout::RowMajor, float, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::ColumnMajor, half_t, layout::RowMajor, half_t, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::RowMajor, half_t, layout::ColumnMajor, float, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::RowMajor, half_t, layout::ColumnMajor, half_t, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::RowMajor, half_t, layout::RowMajor, float, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::RowMajor, half_t, layout::RowMajor, half_t, layout::RowMajor, OpMultiplyAdd > , cutlass::gemm::device::Gemm< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, SplitKSerial, Operator_, IsBetaZero > , cutlass::gemm::device::Gemm< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, SplitKSerial, Operator_, IsBetaZero > , cutlass::gemm::device::GemmBatched< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, Operator_ > , cutlass::gemm::device::GemmBatched< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, Operator_ > , cutlass::gemm::device::GemmComplex< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, TransformA, TransformB, SplitKSerial > , cutlass::gemm::device::GemmComplex< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, TransformA, TransformB, SplitKSerial > , cutlass::gemm::device::GemmSplitKParallel< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ConvertScaledOp_, ReductionOp_, ThreadblockSwizzle_, Stages, kAlignmentA, kAlignmentB, Operator_ > , cutlass::gemm::device::GemmSplitKParallel< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ConvertScaledOp_, ReductionOp_, ThreadblockSwizzle_, Stages, kAlignmentA, kAlignmentB, Operator_ > , cutlass::gemm::kernel::DefaultGemm< int8_t, LayoutA, kAlignmentA, int8_t, LayoutB, kAlignmentB, ElementC, LayoutC, ElementAccumulator, arch::OpClassSimt, ArchTag, ThreadblockShape, WarpShape, GemmShape< 1, 1, 4 >, EpilogueOutputOp, ThreadblockSwizzle, 2, SplitKSerial, Operator, false > , cutlass::gemm::kernel::DefaultGemv< ThreadBlockShape_, ThreadShape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementCD_, LayoutCD_, ElementAccumulator_ > , cutlass::gemm::thread::Mma< Shape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, arch::OpMultiplyAdd, bool > , cutlass::gemm::thread::Mma< Shape_, half_t, LayoutA, half_t, LayoutB, half_t, LayoutC, arch::OpMultiplyAdd > , cutlass::gemm::thread::Mma< Shape_, half_t, LayoutA_, half_t, LayoutB_, half_t, layout::RowMajor, arch::OpMultiplyAdd, typename platform::enable_if< detail::EnableMma_Crow_SM60< LayoutA_, LayoutB_ >::value >::type > , cutlass::gemm::thread::Mma< Shape_, int8_t, layout::ColumnMajor, int8_t, layout::RowMajor, int32_t, LayoutC_, arch::OpMultiplyAdd, int8_t > , cutlass::gemm::thread::Mma< Shape_, int8_t, layout::RowMajor, int8_t, layout::ColumnMajor, int32_t, LayoutC_, arch::OpMultiplyAdd, bool > , cutlass::gemm::thread::MmaGeneric< Shape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, Operator_ > , cutlass::gemm::threadblock::DefaultGemvCore< Shape_, ThreadShape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_ > , cutlass::gemm::threadblock::DefaultMma< int8_t, LayoutA, kAlignmentA, int8_t, LayoutB, kAlignmentB, ElementAccumulator, layout::RowMajor, arch::OpClassSimt, ArchTag, ThreadblockShape, WarpShape, GemmShape< 1, 1, 4 >, 2, Operator, false > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 1 >, ElementA_, layout::ColumnMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 1 >, ElementA_, layout::ColumnMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 1 >, ElementA_, layout::ColumnMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_, > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 1 >, ElementA_, layout::RowMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 1 >, ElementA_, layout::RowMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 4 >, int8_t, layout::ColumnMajor, int8_t, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 4 >, int8_t, layout::ColumnMajor, int8_t, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 4 >, int8_t, layout::RowMajor, int8_t, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 4 >, int8_t, layout::RowMajor, int8_t, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 8, 8, 4 >, ElementA_, layout::ColumnMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 8, 8, 4 >, ElementA_, layout::ColumnMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 8, 8, 4 >, ElementA_, layout::RowMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 8, 8, 4 >, ElementA_, layout::RowMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, InstructionShape_, ElementA_, layout::ColumnMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, InstructionShape_, ElementA_, layout::ColumnMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, InstructionShape_, ElementA_, layout::ColumnMajorInterleaved< InterleavedK >, ElementB_, layout::RowMajorInterleaved< InterleavedK >, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_, AccumulatorsInRowMajor > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, InstructionShape_, ElementA_, layout::RowMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, InstructionShape_, ElementA_, layout::RowMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::warp::MmaComplexTensorOp< Shape_, complex< RealElementA >, LayoutA_, complex< RealElementB >, LayoutB_, complex< RealElementC >, LayoutC_, Policy_, TransformA, TransformB, Enable > , cutlass::gemm::warp::MmaSimt< Shape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, Policy_, PartitionsK, Enable > , cutlass::gemm::warp::MmaTensorOp< Shape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, Policy_, PartitionsK_, AccumulatorsInRowMajor, PartitionsN_, Enable > , cutlass::gemm::warp::MmaVoltaTensorOp< Shape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, Policy_, Enable > , cutlass::reference::device::thread::Gemm< TensorRefA, TensorRefB, TensorRefC, ScalarType, AccumulatorType, OutputTile, InnerProductOp, ConvertOp >
• ElementC : cutlass::arch::Mma< gemm::GemmShape< 16, 8, 8 >, 32, half_t, layout::RowMajor, half_t, layout::ColumnMajor, float, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 16, 8, 8 >, 32, half_t, layout::RowMajor, half_t, layout::ColumnMajor, half_t, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 128 >, 32, uint1b_t, layout::RowMajor, uint1b_t, layout::ColumnMajor, int, layout::RowMajor, OpXorPopc > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, int8_t, layout::RowMajor, int8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, int8_t, layout::RowMajor, int8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, int8_t, layout::RowMajor, uint8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, int8_t, layout::RowMajor, uint8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, uint8_t, layout::RowMajor, int8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, uint8_t, layout::RowMajor, int8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, uint8_t, layout::RowMajor, uint8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 16 >, 32, uint8_t, layout::RowMajor, uint8_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, int4b_t, layout::RowMajor, int4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, int4b_t, layout::RowMajor, int4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, int4b_t, layout::RowMajor, uint4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, int4b_t, layout::RowMajor, uint4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, uint4b_t, layout::RowMajor, int4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, uint4b_t, layout::RowMajor, int4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, uint4b_t, layout::RowMajor, uint4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 32 >, 32, uint4b_t, layout::RowMajor, uint4b_t, layout::ColumnMajor, int, layout::RowMajor, OpMultiplyAddSaturate > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::ColumnMajor, half_t, layout::ColumnMajor, float, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::ColumnMajor, half_t, layout::ColumnMajor, half_t, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::ColumnMajor, half_t, layout::RowMajor, float, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::ColumnMajor, half_t, layout::RowMajor, half_t, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::RowMajor, half_t, layout::ColumnMajor, float, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::RowMajor, half_t, layout::ColumnMajor, half_t, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::RowMajor, half_t, layout::RowMajor, float, layout::RowMajor, OpMultiplyAdd > , cutlass::arch::Mma< gemm::GemmShape< 8, 8, 4 >, 8, half_t, layout::RowMajor, half_t, layout::RowMajor, half_t, layout::RowMajor, OpMultiplyAdd > , cutlass::epilogue::EpilogueWorkspace< Shape_, WarpCount, FragmentC_ > , cutlass::epilogue::warp::FragmentIteratorVoltaTensorOp< WarpShape_, gemm::GemmShape< 32, 32, 4 >, float, layout::RowMajor > , cutlass::epilogue::warp::FragmentIteratorVoltaTensorOp< WarpShape_, gemm::GemmShape< 32, 32, 4 >, half_t, layout::RowMajor > , cutlass::epilogue::warp::VoltaTensorOpPolicy< WarpShape_, gemm::GemmShape< 32, 32, 4 >, float, layout::RowMajor > , cutlass::epilogue::warp::VoltaTensorOpPolicy< WarpShape_, gemm::GemmShape< 32, 32, 4 >, half_t, layout::RowMajor > , cutlass::gemm::device::Gemm< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, SplitKSerial, Operator_, IsBetaZero > , cutlass::gemm::device::Gemm< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, SplitKSerial, Operator_, IsBetaZero > , cutlass::gemm::device::GemmBatched< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, Operator_ > , cutlass::gemm::device::GemmBatched< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, Operator_ > , cutlass::gemm::device::GemmComplex< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, TransformA, TransformB, SplitKSerial > , cutlass::gemm::device::GemmComplex< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, TransformA, TransformB, SplitKSerial > , cutlass::gemm::device::GemmSplitKParallel< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ConvertScaledOp_, ReductionOp_, ThreadblockSwizzle_, Stages, kAlignmentA, kAlignmentB, Operator_ > , cutlass::gemm::device::GemmSplitKParallel< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ConvertScaledOp_, ReductionOp_, ThreadblockSwizzle_, Stages, kAlignmentA, kAlignmentB, Operator_ > , cutlass::gemm::thread::Mma< Shape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, arch::OpMultiplyAdd, bool > , cutlass::gemm::thread::Mma< Shape_, half_t, LayoutA, half_t, LayoutB, half_t, LayoutC, arch::OpMultiplyAdd > , cutlass::gemm::thread::Mma< Shape_, half_t, LayoutA_, half_t, LayoutB_, half_t, layout::RowMajor, arch::OpMultiplyAdd, typename platform::enable_if< detail::EnableMma_Crow_SM60< LayoutA_, LayoutB_ >::value >::type > , cutlass::gemm::thread::Mma< Shape_, int8_t, layout::ColumnMajor, int8_t, layout::RowMajor, int32_t, LayoutC_, arch::OpMultiplyAdd, int8_t > , cutlass::gemm::thread::Mma< Shape_, int8_t, layout::RowMajor, int8_t, layout::ColumnMajor, int32_t, LayoutC_, arch::OpMultiplyAdd, bool > , cutlass::gemm::thread::MmaGeneric< Shape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, Operator_ > , cutlass::gemm::threadblock::DefaultGemvCore< Shape_, ThreadShape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 1 >, ElementA_, layout::ColumnMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 1 >, ElementA_, layout::ColumnMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 1 >, ElementA_, layout::ColumnMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_, > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 1 >, ElementA_, layout::RowMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 1 >, ElementA_, layout::RowMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 4 >, int8_t, layout::ColumnMajor, int8_t, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 4 >, int8_t, layout::ColumnMajor, int8_t, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 4 >, int8_t, layout::RowMajor, int8_t, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 1, 1, 4 >, int8_t, layout::RowMajor, int8_t, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassSimt, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 8, 8, 4 >, ElementA_, layout::ColumnMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 8, 8, 4 >, ElementA_, layout::ColumnMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 8, 8, 4 >, ElementA_, layout::RowMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, GemmShape< 8, 8, 4 >, ElementA_, layout::RowMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, InstructionShape_, ElementA_, layout::ColumnMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, InstructionShape_, ElementA_, layout::ColumnMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, InstructionShape_, ElementA_, layout::ColumnMajorInterleaved< InterleavedK >, ElementB_, layout::RowMajorInterleaved< InterleavedK >, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_, AccumulatorsInRowMajor > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, InstructionShape_, ElementA_, layout::RowMajor, ElementB_, layout::ColumnMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::DefaultMmaCore< Shape_, WarpShape_, InstructionShape_, ElementA_, layout::RowMajor, ElementB_, layout::RowMajor, ElementC_, LayoutC_, arch::OpClassTensorOp, 2, Operator_ > , cutlass::gemm::threadblock::MmaPipelined< Shape_, IteratorA_, SmemIteratorA_, IteratorB_, SmemIteratorB_, ElementC_, LayoutC_, Policy_, TransformA_, TransformB_, Enable > , cutlass::gemm::threadblock::MmaSingleStage< Shape_, IteratorA_, SmemIteratorA_, IteratorB_, SmemIteratorB_, ElementC_, LayoutC_, Policy_, Enable > , cutlass::gemm::warp::MmaComplexTensorOp< Shape_, complex< RealElementA >, LayoutA_, complex< RealElementB >, LayoutB_, complex< RealElementC >, LayoutC_, Policy_, TransformA, TransformB, Enable > , cutlass::gemm::warp::MmaSimt< Shape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, Policy_, PartitionsK, Enable > , cutlass::gemm::warp::MmaTensorOp< Shape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, Policy_, PartitionsK_, AccumulatorsInRowMajor, PartitionsN_, Enable > , cutlass::gemm::warp::MmaVoltaTensorOp< Shape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, Policy_, Enable > , cutlass::reference::device::thread::Gemm< TensorRefA, TensorRefB, TensorRefC, ScalarType, AccumulatorType, OutputTile, InnerProductOp, ConvertOp >
• ElementCD : cutlass::gemm::kernel::DefaultGemv< ThreadBlockShape_, ThreadShape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementCD_, LayoutCD_, ElementAccumulator_ >
• ElementCompute : cutlass::epilogue::thread::Convert< ElementOutput_, Count, ElementAccumulator_, Round > , cutlass::epilogue::thread::LinearCombination< ElementOutput_, Count, ElementAccumulator_, ElementCompute_, Round > , cutlass::epilogue::thread::LinearCombinationClamp< ElementOutput_, Count, ElementAccumulator_, ElementCompute_, Round > , cutlass::epilogue::thread::LinearCombinationRelu< ElementOutput_, Count, ElementAccumulator_, ElementCompute_, Round > , cutlass::epilogue::thread::LinearCombinationRelu< ElementOutput_, Count, int, float, Round >
• ElementOutput : cutlass::epilogue::thread::Convert< ElementOutput_, Count, ElementAccumulator_, Round > , cutlass::epilogue::thread::LinearCombination< ElementOutput_, Count, ElementAccumulator_, ElementCompute_, Round > , cutlass::epilogue::thread::LinearCombinationClamp< ElementOutput_, Count, ElementAccumulator_, ElementCompute_, Round > , cutlass::epilogue::thread::LinearCombinationRelu< ElementOutput_, Count, ElementAccumulator_, ElementCompute_, Round > , cutlass::epilogue::thread::LinearCombinationRelu< ElementOutput_, Count, int, float, Round > , cutlass::epilogue::threadblock::DefaultEpilogueComplexTensorOp< Shape_, WarpMmaTensorOp_, PartitionsK, OutputOp_, ElementsPerAccess > , cutlass::epilogue::threadblock::DefaultEpilogueSimt< Shape_, WarpMmaSimt_, OutputOp_, ElementsPerAccess > , cutlass::epilogue::threadblock::DefaultEpilogueTensorOp< Shape_, WarpMmaTensorOp_, PartitionsK, OutputOp_, ElementsPerAccess > , cutlass::epilogue::threadblock::DefaultEpilogueVoltaTensorOp< Shape_, WarpMmaTensorOp_, PartitionsK, OutputOp_, ElementsPerAccess > , cutlass::epilogue::threadblock::DefaultEpilogueWmmaTensorOp< Shape_, WarpMmaTensorOp_, PartitionsK, OutputOp_, ElementsPerAccess > , cutlass::epilogue::threadblock::DefaultInterleavedEpilogueTensorOp< Shape_, WarpMmaTensorOp_, PartitionsK, OutputOp_, ElementsPerAccess, InterleavedK, IsBetaZero, isSplitK > , cutlass::epilogue::threadblock::DefaultThreadMapVoltaTensorOp< ThreadblockShape_, WarpShape_, PartitionsK, ElementOutput_, ElementsPerAccess, float > , cutlass::epilogue::threadblock::DefaultThreadMapVoltaTensorOp< ThreadblockShape_, WarpShape_, PartitionsK, ElementOutput_, ElementsPerAccess, half_t > , cutlass::epilogue::threadblock::Epilogue< Shape_, WarpMmaOperator_, PartitionsK, OutputTileIterator_, AccumulatorFragmentIterator_, WarpTileIterator_, SharedLoadIterator_, OutputOp_, Padding_ > , cutlass::epilogue::threadblock::InterleavedEpilogue< Shape_, WarpMmaOperator_, PartitionsK, OutputTileIterator_, AccumulatorFragmentIterator_, OutputOp_, InterleavedK, IsBetaZero > , cutlass::reduction::kernel::ReduceSplitK< Shape_, OutputOp_, ReductionOp_, PartitionsPerStage >
• ElementWorkspace : cutlass::reduction::kernel::ReduceSplitK< Shape_, OutputOp_, ReductionOp_, PartitionsPerStage >
• Epilogue : cutlass::epilogue::threadblock::DefaultEpilogueComplexTensorOp< Shape_, WarpMmaTensorOp_, PartitionsK, OutputOp_, ElementsPerAccess > , cutlass::epilogue::threadblock::DefaultEpilogueSimt< Shape_, WarpMmaSimt_, OutputOp_, ElementsPerAccess > , cutlass::epilogue::threadblock::DefaultEpilogueTensorOp< Shape_, WarpMmaTensorOp_, PartitionsK, OutputOp_, ElementsPerAccess > , cutlass::epilogue::threadblock::DefaultEpilogueVoltaTensorOp< Shape_, WarpMmaTensorOp_, PartitionsK, OutputOp_, ElementsPerAccess > , cutlass::epilogue::threadblock::DefaultEpilogueWmmaTensorOp< Shape_, WarpMmaTensorOp_, PartitionsK, OutputOp_, ElementsPerAccess > , cutlass::epilogue::threadblock::DefaultInterleavedEpilogueTensorOp< Shape_, WarpMmaTensorOp_, PartitionsK, OutputOp_, ElementsPerAccess, InterleavedK, IsBetaZero, isSplitK > , cutlass::gemm::kernel::DefaultGemm< ElementA, layout::ColumnMajorInterleaved< InterleavedK >, kAlignmentA, ElementB, layout::RowMajorInterleaved< InterleavedK >, kAlignmentB, ElementC, layout::ColumnMajorInterleaved< InterleavedK >, int32_t, arch::OpClassTensorOp, arch::Sm75, ThreadblockShape, WarpShape, InstructionShape, EpilogueOutputOp, ThreadblockSwizzle, 2, SplitKSerial, Operator, IsBetaZero > , cutlass::gemm::kernel::DefaultGemm< ElementA, LayoutA, kAlignmentA, ElementB, LayoutB, kAlignmentB, ElementC, layout::RowMajor, ElementAccumulator, arch::OpClassSimt, ArchTag, ThreadblockShape, WarpShape, GemmShape< 1, 1, 1 >, EpilogueOutputOp, ThreadblockSwizzle, 2, SplitKSerial, Operator > , cutlass::gemm::kernel::DefaultGemm< ElementA, LayoutA, kAlignmentA, ElementB, LayoutB, kAlignmentB, ElementC, layout::RowMajor, ElementAccumulator, arch::OpClassTensorOp, arch::Sm70, ThreadblockShape, WarpShape, GemmShape< 8, 8, 4 >, EpilogueOutputOp, ThreadblockSwizzle, 2, SplitKSerial, Operator > , cutlass::gemm::kernel::DefaultGemm< ElementA, LayoutA, kAlignmentA, ElementB, LayoutB, kAlignmentB, ElementC, layout::RowMajor, ElementAccumulator, arch::OpClassTensorOp, arch::Sm75, ThreadblockShape, WarpShape, InstructionShape, EpilogueOutputOp, ThreadblockSwizzle, 2, SplitKSerial, Operator > , cutlass::gemm::kernel::DefaultGemm< int8_t, LayoutA, kAlignmentA, int8_t, LayoutB, kAlignmentB, ElementC, LayoutC, ElementAccumulator, arch::OpClassSimt, ArchTag, ThreadblockShape, WarpShape, GemmShape< 1, 1, 4 >, EpilogueOutputOp, ThreadblockSwizzle, 2, SplitKSerial, Operator, false > , cutlass::gemm::kernel::DefaultGemmSplitKParallel< ElementA_, LayoutA_, kAlignmentA, ElementB_, LayoutB_, kAlignmentB, ElementC_, LayoutC_, ElementAccumulator, OperatorClass, ArchTag, ThreadblockShape, WarpShape, InstructionShape, EpilogueOutputOp, ThreadblockSwizzle, Stages, Operator > , cutlass::gemm::kernel::Gemm< Mma_, Epilogue_, ThreadblockSwizzle_, SplitKSerial > , cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ > , cutlass::gemm::kernel::GemmSplitKParallel< Mma_, Epilogue_, ThreadblockSwizzle_ >
• EpilogueOutputOp : cutlass::gemm::device::DefaultGemmConfiguration< arch::OpClassSimt, ArchTag, ElementA, ElementB, ElementC, ElementAccumulator > , cutlass::gemm::device::DefaultGemmConfiguration< arch::OpClassSimt, ArchTag, int8_t, int8_t, ElementC, int32_t > , cutlass::gemm::device::DefaultGemmConfiguration< arch::OpClassTensorOp, arch::Sm70, ElementA, ElementB, ElementC, ElementAccumulator > , cutlass::gemm::device::DefaultGemmConfiguration< arch::OpClassTensorOp, arch::Sm75, ElementA, ElementB, ElementC, ElementAccumulator > , cutlass::gemm::device::DefaultGemmConfiguration< arch::OpClassTensorOp, arch::Sm75, int4b_t, int4b_t, ElementC, int32_t > , cutlass::gemm::device::DefaultGemmConfiguration< arch::OpClassTensorOp, arch::Sm75, int4b_t, uint4b_t, ElementC, int32_t > , cutlass::gemm::device::DefaultGemmConfiguration< arch::OpClassTensorOp, arch::Sm75, int8_t, int8_t, ElementC, int32_t > , cutlass::gemm::device::DefaultGemmConfiguration< arch::OpClassTensorOp, arch::Sm75, int8_t, uint8_t, ElementC, int32_t > , cutlass::gemm::device::DefaultGemmConfiguration< arch::OpClassTensorOp, arch::Sm75, uint4b_t, int4b_t, ElementC, int32_t > , cutlass::gemm::device::DefaultGemmConfiguration< arch::OpClassTensorOp, arch::Sm75, uint4b_t, uint4b_t, ElementC, int32_t > , cutlass::gemm::device::DefaultGemmConfiguration< arch::OpClassTensorOp, arch::Sm75, uint8_t, int8_t, ElementC, int32_t > , cutlass::gemm::device::DefaultGemmConfiguration< arch::OpClassTensorOp, arch::Sm75, uint8_t, uint8_t, ElementC, int32_t > , cutlass::gemm::device::DefaultGemmConfiguration< arch::OpClassWmmaTensorOp, ArchTag, ElementA, ElementB, ElementC, ElementAccumulator > , cutlass::gemm::device::Gemm< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, SplitKSerial, Operator_, IsBetaZero > , cutlass::gemm::device::Gemm< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, SplitKSerial, Operator_, IsBetaZero > , cutlass::gemm::device::GemmBatched< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, Operator_ > , cutlass::gemm::device::GemmBatched< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, AlignmentA, AlignmentB, Operator_ > , cutlass::gemm::device::GemmComplex< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, TransformA, TransformB, SplitKSerial > , cutlass::gemm::device::GemmComplex< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ThreadblockSwizzle_, Stages, TransformA, TransformB, SplitKSerial > , cutlass::gemm::device::GemmSplitKParallel< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, LayoutC_, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ConvertScaledOp_, ReductionOp_, ThreadblockSwizzle_, Stages, kAlignmentA, kAlignmentB, Operator_ > , cutlass::gemm::device::GemmSplitKParallel< ElementA_, LayoutA_, ElementB_, LayoutB_, ElementC_, layout::ColumnMajor, ElementAccumulator_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, EpilogueOutputOp_, ConvertScaledOp_, ReductionOp_, ThreadblockSwizzle_, Stages, kAlignmentA, kAlignmentB, Operator_ >