- f -

• FloatType : cutlass::reference::device::detail::RandomGaussianFunc< Element > , cutlass::reference::device::detail::RandomUniformFunc< Element >
• Fragment : cutlass::epilogue::thread::ReductionOpPlus< Element_, Count > , cutlass::epilogue::threadblock::InterleavedPredicatedTileIterator< ThreadMap_, Element_, InterleavedK > , cutlass::epilogue::threadblock::PredicatedTileIterator< ThreadMap_, Element_ > , cutlass::epilogue::threadblock::SharedLoadIterator< ThreadMap_, Element_, MaxAlignment > , cutlass::epilogue::warp::FragmentIteratorComplexTensorOp< WarpShape_, OperatorShape_, OperatorElementC_, OperatorFragmentC_, layout::RowMajor > , cutlass::epilogue::warp::FragmentIteratorSimt< WarpShape_, Operator_, layout::RowMajor, MmaSimtPolicy_ > , cutlass::epilogue::warp::FragmentIteratorTensorOp< WarpShape_, OperatorShape_, OperatorElementC_, OperatorFragmentC_, layout::ColumnMajorInterleaved< InterleavedK > > , cutlass::epilogue::warp::FragmentIteratorTensorOp< WarpShape_, OperatorShape_, OperatorElementC_, OperatorFragmentC_, layout::RowMajor > , 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::FragmentIteratorWmmaTensorOp< WarpShape_, OperatorShape_, OperatorElementC_, OperatorFragmentC_, layout::RowMajor > , 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::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::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::transform::thread::Transpose< ElementCount_, layout::PitchLinearShape< 4, 4 >, int8_t > , 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::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 >
• FragmentA : 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::thread::detail::Mma_HFMA2< Shape, layout::ColumnMajor, layout::ColumnMajor, layout::ColumnMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::ColumnMajor, layout::ColumnMajor, layout::RowMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::ColumnMajor, layout::RowMajor, layout::ColumnMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::ColumnMajor, layout::RowMajor, layout::RowMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::RowMajor, layout::ColumnMajor, layout::ColumnMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::RowMajor, layout::ColumnMajor, layout::RowMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::RowMajor, layout::RowMajor, layout::ColumnMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::RowMajor, layout::RowMajor, layout::RowMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, LayoutA, LayoutB, layout::ColumnMajor, false > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, LayoutA, LayoutB, layout::RowMajor, false > , 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::Gemv< Core_ > , 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 >
• FragmentAccumulator : 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::reduction::kernel::ReduceSplitK< Shape_, OutputOp_, ReductionOp_, PartitionsPerStage > , cutlass::reduction::thread::ReduceAdd< ElementAccumulator_, Element_, Count >
• FragmentB : 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::thread::detail::Mma_HFMA2< Shape, layout::ColumnMajor, layout::ColumnMajor, layout::ColumnMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::ColumnMajor, layout::ColumnMajor, layout::RowMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::ColumnMajor, layout::RowMajor, layout::ColumnMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::ColumnMajor, layout::RowMajor, layout::RowMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::RowMajor, layout::ColumnMajor, layout::ColumnMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::RowMajor, layout::ColumnMajor, layout::RowMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::RowMajor, layout::RowMajor, layout::ColumnMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::RowMajor, layout::RowMajor, layout::RowMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, LayoutA, LayoutB, layout::ColumnMajor, false > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, LayoutA, LayoutB, layout::RowMajor, false > , 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::Gemv< Core_ > , 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 >
• FragmentC : 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::threadblock::DirectEpilogueTensorOp< Shape_, Operator_, PartitionsK, Element_, OutputOp_, ConvertOp_ > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::ColumnMajor, layout::ColumnMajor, layout::ColumnMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::ColumnMajor, layout::ColumnMajor, layout::RowMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::ColumnMajor, layout::RowMajor, layout::ColumnMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::ColumnMajor, layout::RowMajor, layout::RowMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::RowMajor, layout::ColumnMajor, layout::ColumnMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::RowMajor, layout::ColumnMajor, layout::RowMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::RowMajor, layout::RowMajor, layout::ColumnMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, layout::RowMajor, layout::RowMajor, layout::RowMajor, true > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, LayoutA, LayoutB, layout::ColumnMajor, false > , cutlass::gemm::thread::detail::Mma_HFMA2< Shape, LayoutA, LayoutB, layout::RowMajor, false > , 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::Gemv< Core_ > , 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 >
• FragmentCD : cutlass::gemm::kernel::DefaultGemv< ThreadBlockShape_, ThreadShape_, ElementA_, LayoutA_, ElementB_, LayoutB_, ElementCD_, LayoutCD_, ElementAccumulator_ >
• FragmentElement : cutlass::reduction::thread::ReduceAdd< ElementAccumulator_, Element_, Count >
• FragmentOutput : 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::reduction::kernel::ReduceSplitK< Shape_, OutputOp_, ReductionOp_, PartitionsPerStage >
• FragmentWorkspace : cutlass::reduction::kernel::ReduceSplitK< Shape_, OutputOp_, ReductionOp_, PartitionsPerStage >
• Functor : cutlass::reduction::BatchedReduction< BatchedReductionTraits_ > , cutlass::reduction::BatchedReductionTraits< ScalarA_, ScalarC_, ScalarD_, ScalarAlphaBeta_, ScalarAccum_, ReductionSize_, OutputTile_, SubTile_, ThreadShape_, Index_, BlockSwizzle_, maxInReg_, maxOutReg_, Functor_ >