CUTLASS
CUDA Templates for Linear Algebra Subroutines and Solvers
gemm/thread/mma.h
Go to the documentation of this file.
1 /***************************************************************************************************
2  * Copyright (c) 2017-2019, NVIDIA CORPORATION. All rights reserved.
3  *
4  * Redistribution and use in source and binary forms, with or without modification, are permitted
5  * provided that the following conditions are met:
6  * * Redistributions of source code must retain the above copyright notice, this list of
7  * conditions and the following disclaimer.
8  * * Redistributions in binary form must reproduce the above copyright notice, this list of
9  * conditions and the following disclaimer in the documentation and/or other materials
10  * provided with the distribution.
11  * * Neither the name of the NVIDIA CORPORATION nor the names of its contributors may be used
12  * to endorse or promote products derived from this software without specific prior written
13  * permission.
14  *
15  * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR
16  * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
17  * FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL NVIDIA CORPORATION BE LIABLE
18  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
19  * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
20  * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
21  * STRICT LIABILITY, OR TOR (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
22  * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
23  *
24  **************************************************************************************************/
29 #pragma once
30 
31 #include "cutlass/cutlass.h"
32 #include "cutlass/array.h"
33 #include "cutlass/numeric_types.h"
34 #include "cutlass/gemm/gemm.h"
35 #include "cutlass/arch/mma.h"
36 
38 
39 namespace cutlass {
40 namespace gemm {
41 namespace thread {
42 
44 
46 template <
48  typename Shape,
50  typename ElementA,
52  typename LayoutA,
54  typename ElementB,
56  typename LayoutB,
58  typename ElementC,
60  typename LayoutC,
62  typename Operator = arch::OpMultiplyAdd,
64  typename Enable = bool
65 >
66 struct Mma;
67 
69 
70 } // namespace thread
71 } // namespace gemm
72 } // namespace cutlass
73 
75 
76 //
77 // Overloads specialized for existing architectures
78 //
79 
83 
Definition: aligned_buffer.h:35
Templates exposing architecture support for multiply-add operations.
Defines common types used for all GEMM-like operators.
Statically sized array of elements that accommodates all CUTLASS-supported numeric types and is safe ...
Templates exposing architecture support for multiply-add operations.
Templates exposing architecture support for multiply-add operations.
Templates exposing architecture support for multiply-add operations.
Top-level include for all CUTLASS numeric types.
Structure to compute the matrix product.
Definition: gemm/thread/mma.h:66
Basic include for CUTLASS.