Template for a pipelined GEMM kernel. Does not compute batching or support split-K.
More...
|
struct | cutlass::gemm::threadblock::DefaultMma< ElementA_, LayoutA_, kAlignmentA, ElementB_, LayoutB_, kAlignmentB, ElementAccumulator_, LayoutC_, OperatorClass_, ArchTag_, ThreadblockShape_, WarpShape_, InstructionShape_, Stages, Operator, AccumulatorsInRowMajor > |
|
struct | cutlass::gemm::threadblock::DefaultMma< ElementA, LayoutA, kAlignmentA, ElementB, LayoutB, kAlignmentB, ElementAccumulator, layout::RowMajor, arch::OpClassSimt, ArchTag, ThreadblockShape, WarpShape, InstructionShape, 2, Operator, false > |
| Specialization for row-major output (OperatorClass Simt) More...
|
|
struct | cutlass::gemm::threadblock::DefaultMma< ElementA, LayoutA, kAlignmentA, ElementB, LayoutB, kAlignmentB, ElementAccumulator, layout::RowMajor, arch::OpClassTensorOp, ArchTag, ThreadblockShape, WarpShape, InstructionShape, 2, Operator, false > |
| Specialization for row-major output (OperatorClass Simt) More...
|
|
struct | cutlass::gemm::threadblock::DefaultMma< ElementA, LayoutA, kAlignmentA, ElementB, LayoutB, kAlignmentB, ElementAccumulator, layout::ColumnMajorInterleaved< InterleavedK >, OperatorClass, ArchTag, ThreadblockShape, WarpShape, InstructionShape, 2, Operator, true > |
| Specialization for column-major-interleaved output. More...
|
|
struct | 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 > |
|