CUTLASS: cutlass::epilogue::threadblock::InterleavedOutputTileThreadMap< WarpCount_, MmaCount_, Threads, ElementsPerAccess, ElementSize > Struct Template Reference

| | CUTLASS

CUDA Templates for Linear Algebra Subroutines and Solvers |

• Main Page

• Modules

• Namespaces

• Classes

• Files

• Class List

• Class Index

• Class Hierarchy

• Class Members

• cutlass

• epilogue

• threadblock

• InterleavedOutputTileThreadMap

Classes | Public Types | Static Public Member Functions | Static Public Attributes | List of all members

cutlass::epilogue::threadblock::InterleavedOutputTileThreadMap< WarpCount_, MmaCount_, Threads, ElementsPerAccess, ElementSize > Struct Template Reference

#include <output_tile_thread_map.h>

|

Classes

| | struct | Detail | | |

|

Public Types

| | using | WarpCount = WarpCount_ | | | | using | MmaCount = MmaCount_ | | | | using | Iterations = MmaCount | | | | using | Delta = layout::PitchLinearShape< kWarpSize *kElementsPerAccess, 1 > | | |

|

Static Public Member Functions

| | static CUTLASS_HOST_DEVICE layout::PitchLinearCoord | initial_offset (int thread_idx) | | | Initial offset function. More...
| | |

|

Static Public Attributes

| | static int const | kWarpSize = 32 | | | | static int const | kThreads = Threads | | | | static int const | kWarpCount = kThreads / kWarpSize | | | | static int const | kElementsPerAccess = ElementsPerAccess | | | | static int const | kElementSize = ElementSize | | |

Detailed Description

template<typename WarpCount_, typename MmaCount_, int Threads, int ElementsPerAccess, int ElementSize> struct cutlass::epilogue::threadblock::InterleavedOutputTileThreadMap< WarpCount_, MmaCount_, Threads, ElementsPerAccess, ElementSize >

Template metaprogram for partitioning a 3D interleaved layout across warps to achieve several performance objectives:

• coalesced memory accesses in units of 64 Byte lines
• minimal address arithmetic
• minimal predicate calculations

Member Typedef Documentation

template<typename WarpCount_ , typename MmaCount_ , int Threads, int ElementsPerAccess, int ElementSize>

| using cutlass::epilogue::threadblock::InterleavedOutputTileThreadMap< WarpCount_, MmaCount_, Threads, ElementsPerAccess, ElementSize >::Delta = layout::PitchLinearShape<kWarpSize * kElementsPerAccess, 1> |

template<typename WarpCount_ , typename MmaCount_ , int Threads, int ElementsPerAccess, int ElementSize>

| using cutlass::epilogue::threadblock::InterleavedOutputTileThreadMap< WarpCount_, MmaCount_, Threads, ElementsPerAccess, ElementSize >::Iterations = MmaCount |

template<typename WarpCount_ , typename MmaCount_ , int Threads, int ElementsPerAccess, int ElementSize>

| using cutlass::epilogue::threadblock::InterleavedOutputTileThreadMap< WarpCount_, MmaCount_, Threads, ElementsPerAccess, ElementSize >::MmaCount = MmaCount_ |

template<typename WarpCount_ , typename MmaCount_ , int Threads, int ElementsPerAccess, int ElementSize>

| using cutlass::epilogue::threadblock::InterleavedOutputTileThreadMap< WarpCount_, MmaCount_, Threads, ElementsPerAccess, ElementSize >::WarpCount = WarpCount_ |

Member Function Documentation

template<typename WarpCount_ , typename MmaCount_ , int Threads, int ElementsPerAccess, int ElementSize>

|

| static CUTLASS_HOST_DEVICE layout::PitchLinearCoord cutlass::epilogue::threadblock::InterleavedOutputTileThreadMap< WarpCount_, MmaCount_, Threads, ElementsPerAccess, ElementSize >::initial_offset | ( | int | thread_idx | ) | |

| inlinestatic |

Member Data Documentation

template<typename WarpCount_ , typename MmaCount_ , int Threads, int ElementsPerAccess, int ElementSize>

|

| int const cutlass::epilogue::threadblock::InterleavedOutputTileThreadMap< WarpCount_, MmaCount_, Threads, ElementsPerAccess, ElementSize >::kElementSize = ElementSize |

| static |

template<typename WarpCount_ , typename MmaCount_ , int Threads, int ElementsPerAccess, int ElementSize>

|

| int const cutlass::epilogue::threadblock::InterleavedOutputTileThreadMap< WarpCount_, MmaCount_, Threads, ElementsPerAccess, ElementSize >::kElementsPerAccess = ElementsPerAccess |

| static |

template<typename WarpCount_ , typename MmaCount_ , int Threads, int ElementsPerAccess, int ElementSize>

|

| int const cutlass::epilogue::threadblock::InterleavedOutputTileThreadMap< WarpCount_, MmaCount_, Threads, ElementsPerAccess, ElementSize >::kThreads = Threads |

| static |

template<typename WarpCount_ , typename MmaCount_ , int Threads, int ElementsPerAccess, int ElementSize>

|

| int const cutlass::epilogue::threadblock::InterleavedOutputTileThreadMap< WarpCount_, MmaCount_, Threads, ElementsPerAccess, ElementSize >::kWarpCount = kThreads / kWarpSize |

| static |

template<typename WarpCount_ , typename MmaCount_ , int Threads, int ElementsPerAccess, int ElementSize>

|

| int const cutlass::epilogue::threadblock::InterleavedOutputTileThreadMap< WarpCount_, MmaCount_, Threads, ElementsPerAccess, ElementSize >::kWarpSize = 32 |

| static |

---

The documentation for this struct was generated from the following file:

• [output_tile_thread_map.h](output tile thread__map_8h_source.html)

---

Generated by 1.8.11