158 lines
99 KiB
HTML
158 lines
99 KiB
HTML
|
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
|
||
|
|
<html xmlns="http://www.w3.org/1999/xhtml">
|
||
|
|
<head>
|
||
|
|
<meta http-equiv="Content-Type" content="text/xhtml;charset=UTF-8"/>
|
||
|
|
<meta http-equiv="X-UA-Compatible" content="IE=9"/>
|
||
|
|
<meta name="generator" content="Doxygen 1.8.11"/>
|
||
|
|
<title>CUTLASS: default_gemm.h Source File</title>
|
||
|
|
<link href="tabs.css" rel="stylesheet" type="text/css"/>
|
||
|
|
<script type="text/javascript" src="jquery.js"></script>
|
||
|
|
<script type="text/javascript" src="dynsections.js"></script>
|
||
|
|
<link href="search/search.css" rel="stylesheet" type="text/css"/>
|
||
|
|
<script type="text/javascript" src="search/searchdata.js"></script>
|
||
|
|
<script type="text/javascript" src="search/search.js"></script>
|
||
|
|
<script type="text/javascript">
|
||
|
|
$(document).ready(function() { init_search(); });
|
||
|
|
</script>
|
||
|
|
<script type="text/x-mathjax-config">
|
||
|
|
MathJax.Hub.Config({
|
||
|
|
extensions: ["tex2jax.js"],
|
||
|
|
jax: ["input/TeX","output/HTML-CSS"],
|
||
|
|
});
|
||
|
|
</script><script type="text/javascript" src="http://cdn.mathjax.org/mathjax/latest/MathJax.js"></script>
|
||
|
|
<link href="doxygen.css" rel="stylesheet" type="text/css" />
|
||
|
|
</head>
|
||
|
|
<body>
|
||
|
|
<div id="top"><!-- do not remove this div, it is closed by doxygen! -->
|
||
|
|
<div id="titlearea">
|
||
|
|
<table cellspacing="0" cellpadding="0">
|
||
|
|
<tbody>
|
||
|
|
<tr style="height: 56px;">
|
||
|
|
<td id="projectlogo"><img alt="Logo" src="cutlass-logo-small.png"/></td>
|
||
|
|
<td id="projectalign" style="padding-left: 0.5em;">
|
||
|
|
<div id="projectname">CUTLASS
|
||
|
|
</div>
|
||
|
|
<div id="projectbrief">CUDA Templates for Linear Algebra Subroutines and Solvers</div>
|
||
|
|
</td>
|
||
|
|
</tr>
|
||
|
|
</tbody>
|
||
|
|
</table>
|
||
|
|
</div>
|
||
|
|
<!-- end header part -->
|
||
|
|
<!-- Generated by Doxygen 1.8.11 -->
|
||
|
|
<script type="text/javascript">
|
||
|
|
var searchBox = new SearchBox("searchBox", "search",false,'Search');
|
||
|
|
</script>
|
||
|
|
<div id="navrow1" class="tabs">
|
||
|
|
<ul class="tablist">
|
||
|
|
<li><a href="index.html"><span>Main Page</span></a></li>
|
||
|
|
<li><a href="modules.html"><span>Modules</span></a></li>
|
||
|
|
<li><a href="namespaces.html"><span>Namespaces</span></a></li>
|
||
|
|
<li><a href="annotated.html"><span>Classes</span></a></li>
|
||
|
|
<li class="current"><a href="files.html"><span>Files</span></a></li>
|
||
|
|
<li>
|
||
|
|
<div id="MSearchBox" class="MSearchBoxInactive">
|
||
|
|
<span class="left">
|
||
|
|
<img id="MSearchSelect" src="search/mag_sel.png"
|
||
|
|
onmouseover="return searchBox.OnSearchSelectShow()"
|
||
|
|
onmouseout="return searchBox.OnSearchSelectHide()"
|
||
|
|
alt=""/>
|
||
|
|
<input type="text" id="MSearchField" value="Search" accesskey="S"
|
||
|
|
onfocus="searchBox.OnSearchFieldFocus(true)"
|
||
|
|
onblur="searchBox.OnSearchFieldFocus(false)"
|
||
|
|
onkeyup="searchBox.OnSearchFieldChange(event)"/>
|
||
|
|
</span><span class="right">
|
||
|
|
<a id="MSearchClose" href="javascript:searchBox.CloseResultsWindow()"><img id="MSearchCloseImg" border="0" src="search/close.png" alt=""/></a>
|
||
|
|
</span>
|
||
|
|
</div>
|
||
|
|
</li>
|
||
|
|
</ul>
|
||
|
|
</div>
|
||
|
|
<div id="navrow2" class="tabs2">
|
||
|
|
<ul class="tablist">
|
||
|
|
<li><a href="files.html"><span>File List</span></a></li>
|
||
|
|
<li><a href="globals.html"><span>File Members</span></a></li>
|
||
|
|
</ul>
|
||
|
|
</div>
|
||
|
|
<!-- window showing the filter options -->
|
||
|
|
<div id="MSearchSelectWindow"
|
||
|
|
onmouseover="return searchBox.OnSearchSelectShow()"
|
||
|
|
onmouseout="return searchBox.OnSearchSelectHide()"
|
||
|
|
onkeydown="return searchBox.OnSearchSelectKey(event)">
|
||
|
|
</div>
|
||
|
|
|
||
|
|
<!-- iframe showing the search results (closed by default) -->
|
||
|
|
<div id="MSearchResultsWindow">
|
||
|
|
<iframe src="javascript:void(0)" frameborder="0"
|
||
|
|
name="MSearchResults" id="MSearchResults">
|
||
|
|
</iframe>
|
||
|
|
</div>
|
||
|
|
|
||
|
|
<div id="nav-path" class="navpath">
|
||
|
|
<ul>
|
||
|
|
<li class="navelem"><a class="el" href="dir_d44c64559bbebec7f509842c48db8b23.html">include</a></li><li class="navelem"><a class="el" href="dir_6baf2bb612a2f0daa69af3101ede80a1.html">cutlass</a></li><li class="navelem"><a class="el" href="dir_9aa36bd9cfad59a1f88859a38871c977.html">gemm</a></li><li class="navelem"><a class="el" href="dir_c4a2560cb67fbf4e24d3d775f040b990.html">kernel</a></li> </ul>
|
||
|
|
</div>
|
||
|
|
</div><!-- top -->
|
||
|
|
<div class="header">
|
||
|
|
<div class="headertitle">
|
||
|
|
<div class="title">default_gemm.h</div> </div>
|
||
|
|
</div><!--header-->
|
||
|
|
<div class="contents">
|
||
|
|
<a href="default__gemm_8h.html">Go to the documentation of this file.</a><div class="fragment"><div class="line"><a name="l00001"></a><span class="lineno"> 1</span> <span class="comment">/***************************************************************************************************</span></div><div class="line"><a name="l00002"></a><span class="lineno"> 2</span> <span class="comment"> * Copyright (c) 2017-2019, NVIDIA CORPORATION. All rights reserved.</span></div><div class="line"><a name="l00003"></a><span class="lineno"> 3</span> <span class="comment"> *</span></div><div class="line"><a name="l00004"></a><span class="lineno"> 4</span> <span class="comment"> * Redistribution and use in source and binary forms, with or without modification, are permitted</span></div><div class="line"><a name="l00005"></a><span class="lineno"> 5</span> <span class="comment"> * provided that the following conditions are met:</span></div><div class="line"><a name="l00006"></a><span class="lineno"> 6</span> <span class="comment"> * * Redistributions of source code must retain the above copyright notice, this list of</span></div><div class="line"><a name="l00007"></a><span class="lineno"> 7</span> <span class="comment"> * conditions and the following disclaimer.</span></div><div class="line"><a name="l00008"></a><span class="lineno"> 8</span> <span class="comment"> * * Redistributions in binary form must reproduce the above copyright notice, this list of</span></div><div class="line"><a name="l00009"></a><span class="lineno"> 9</span> <span class="comment"> * conditions and the following disclaimer in the documentation and/or other materials</span></div><div class="line"><a name="l00010"></a><span class="lineno"> 10</span> <span class="comment"> * provided with the distribution.</span></div><div class="line"><a name="l00011"></a><span class="lineno"> 11</span> <span class="comment"> * * Neither the name of the NVIDIA CORPORATION nor the names of its contributors may be used</span></div><div class="line"><a name="l00012"></a><span class="lineno"> 12</span> <span class="comment"> * to endorse or promote products derived from this software without specific prior written</span></div><div class="line"><a name="l00013"></a><span class="lineno"> 13</span> <span class="comment"> * permission.</span></div><div class="line"><a name="l00014"></a><span class="lineno"> 14</span> <span class="comment"> *</span></div><div class="line"><a name="l00015"></a><span class="lineno"> 15</span> <span class="comment"> * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR</span></div><div class="line"><a name="l00016"></a><span class="lineno"> 16</span> <span class="comment"> * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND</span></div><div class="line"><a name="l00017"></a><span class="lineno"> 17</span> <span class="comment"> * FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL NVIDIA CORPORATION BE LIABLE</span></div><div class="line"><a name="l00018"></a><span class="lineno"> 18</span> <span class="comment"> * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,</span></div><div class="line"><a name="l00019"></a><span class="lineno"> 19</span> <span class="comment"> * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;</span></div><div class="line"><a name="l00020"></a><span class="lineno"> 20</span> <span class="comment"> * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,</span></div><div class="line"><a name="l00021"></a><span class="lineno"> 21</span> <span class="comment"> * STRICT LIABILITY, OR TOR (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE</span></div><div class="line"><a name="l00022
|
||
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_html"><div class="ttname"><a href="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm.html">cutlass::gemm::kernel::DefaultGemm</a></div><div class="ttdef"><b>Definition:</b> default_gemm.h:116</div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00_01E5d78d37a9ae2ec08d7d477d571df036e_html_a04323cdfbd5109ceabcb5bf31d435798"><div class="ttname"><a href="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00_01E5d78d37a9ae2ec08d7d477d571df036e.html#a04323cdfbd5109ceabcb5bf31d435798">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 >::Mma</a></div><div class="ttdeci">typename cutlass::gemm::threadblock::DefaultMma< ElementA, LayoutA, kAlignmentA, ElementB, LayoutB, kAlignmentB, ElementAccumulator, layout::RowMajor, arch::OpClassTensorOp, arch::Sm75, ThreadblockShape, WarpShape, InstructionShape, 2, Operator >::ThreadblockMma Mma</div><div class="ttdoc">Define the threadblock-scoped matrix multiply-accumulate. </div><div class="ttdef"><b>Definition:</b> default_gemm.h:186</div></div>
|
||
|
|
<div class="ttc" id="namespacecutlass_html"><div class="ttname"><a href="namespacecutlass.html">cutlass</a></div><div class="ttdef"><b>Definition:</b> aligned_buffer.h:35</div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1epilogue_1_1threadblock_1_1DefaultEpilogueSimt_html"><div class="ttname"><a href="structcutlass_1_1epilogue_1_1threadblock_1_1DefaultEpilogueSimt.html">cutlass::epilogue::threadblock::DefaultEpilogueSimt</a></div><div class="ttdoc">Defines sensible defaults for epilogues for SimtOps. </div><div class="ttdef"><b>Definition:</b> default_epilogue_simt.h:70</div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1arch_1_1Sm50_html"><div class="ttname"><a href="structcutlass_1_1arch_1_1Sm50.html">cutlass::arch::Sm50</a></div><div class="ttdef"><b>Definition:</b> arch.h:37</div></div>
|
||
|
|
<div class="ttc" id="default__epilogue__wmma__tensor__op_8h_html"><div class="ttname"><a href="default__epilogue__wmma__tensor__op_8h.html">default_epilogue_wmma_tensor_op.h</a></div><div class="ttdoc">Epilogue for threadblock scoped GEMMs using Tensor Ops. </div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1arch_1_1Sm70_html"><div class="ttname"><a href="structcutlass_1_1arch_1_1Sm70.html">cutlass::arch::Sm70</a></div><div class="ttdef"><b>Definition:</b> arch.h:46</div></div>
|
||
|
|
<div class="ttc" id="include_2cutlass_2gemm_2gemm_8h_html"><div class="ttname"><a href="include_2cutlass_2gemm_2gemm_8h.html">gemm.h</a></div><div class="ttdoc">Defines common types used for all GEMM-like operators. </div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01ElementA_00_01layout_1_1ColumnMajorInterleave661fe54d13cc2c9153dcdf31e4beaa30_html_a3ff721fd6df7783d2481ff466ef9041b"><div class="ttname"><a href="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01ElementA_00_01layout_1_1ColumnMajorInterleave661fe54d13cc2c9153dcdf31e4beaa30.html#a3ff721fd6df7783d2481ff466ef9041b">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 >::Mma</a></div><div class="ttdeci">typename cutlass::gemm::threadblock::DefaultMma< ElementA, LayoutA, kAlignmentA, ElementB, LayoutB, kAlignmentB, ElementAccumulator, LayoutC, arch::OpClassTensorOp, arch::Sm75, ThreadblockShape, WarpShape, InstructionShape, 2, Operator, true >::ThreadblockMma Mma</div><div class="ttdoc">Define the threadblock-scoped matrix multiply-accumulate. </div><div class="ttdef"><b>Definition:</b> default_gemm.h:252</div></div>
|
||
|
|
<div class="ttc" id="default__mma_8h_html"><div class="ttname"><a href="default__mma_8h.html">default_mma.h</a></div><div class="ttdoc">Template for a pipelined GEMM kernel. Does not compute batching or support split-K. </div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_01inf48440732c1c5f42ddbfaba179861815_html_a9c029aa1ac8275a18aeaefa8a4469dfb"><div class="ttname"><a href="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_01inf48440732c1c5f42ddbfaba179861815.html#a9c029aa1ac8275a18aeaefa8a4469dfb">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 >::Mma</a></div><div class="ttdeci">typename cutlass::gemm::threadblock::DefaultMma< ElementA, LayoutA, kAlignmentA, ElementB, LayoutB, kAlignmentB, ElementAccumulator, LayoutC, arch::OpClassSimt, arch::Sm50, ThreadblockShape, WarpShape, InstructionShape, 2, Operator, false >::ThreadblockMma Mma</div><div class="ttdoc">Define the threadblock-scoped matrix multiply-accumulate. </div><div class="ttdef"><b>Definition:</b> default_gemm.h:502</div></div>
|
||
|
|
<div class="ttc" id="default__mma__core__sm70_8h_html"><div class="ttname"><a href="default__mma__core__sm70_8h.html">default_mma_core_sm70.h</a></div><div class="ttdoc">Defines basic properties needed by CTA-level GEMMs assuming expectations about data layout of the glo...</div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_html"><div class="ttname"><a href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma.html">cutlass::gemm::threadblock::DefaultMma</a></div><div class="ttdef"><b>Definition:</b> default_mma.h:87</div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1arch_1_1Sm75_html"><div class="ttname"><a href="structcutlass_1_1arch_1_1Sm75.html">cutlass::arch::Sm75</a></div><div class="ttdef"><b>Definition:</b> arch.h:52</div></div>
|
||
|
|
<div class="ttc" id="linear__combination_8h_html"><div class="ttname"><a href="linear__combination_8h.html">linear_combination.h</a></div><div class="ttdoc">Functor performing linear combination operations used by epilogues. </div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1sizeof__bits_html"><div class="ttname"><a href="structcutlass_1_1sizeof__bits.html">cutlass::sizeof_bits</a></div><div class="ttdoc">Defines the size of an element in bits. </div><div class="ttdef"><b>Definition:</b> numeric_types.h:42</div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00_01E044b039b2fe402f29b04a9f5feee5342_html_a39c63f9ec6ae91ac66ff51f7e132b27d"><div class="ttname"><a href="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00_01E044b039b2fe402f29b04a9f5feee5342.html#a39c63f9ec6ae91ac66ff51f7e132b27d">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 >::Epilogue</a></div><div class="ttdeci">typename cutlass::epilogue::threadblock::DefaultEpilogueVoltaTensorOp< ThreadblockShape, typename Mma::Operator, kPartitionsK, EpilogueOutputOp, EpilogueOutputOp::kCount >::Epilogue Epilogue</div><div class="ttdoc">Define the epilogue. </div><div class="ttdef"><b>Definition:</b> default_gemm.h:346</div></div>
|
||
|
|
<div class="ttc" id="numeric__types_8h_html"><div class="ttname"><a href="numeric__types_8h.html">numeric_types.h</a></div><div class="ttdoc">Top-level include for all CUTLASS numeric types. </div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1GemmShape_html"><div class="ttname"><a href="structcutlass_1_1gemm_1_1GemmShape.html">cutlass::gemm::GemmShape</a></div><div class="ttdoc">Shape of a matrix multiply-add operation. </div><div class="ttdef"><b>Definition:</b> include/cutlass/gemm/gemm.h:57</div></div>
|
||
|
|
<div class="ttc" id="platform_8h_html_adde4c9ea91b753491851361a4198c009"><div class="ttname"><a href="platform_8h.html#adde4c9ea91b753491851361a4198c009">static_assert</a></div><div class="ttdeci">#define static_assert(__e, __m)</div><div class="ttdef"><b>Definition:</b> platform.h:153</div></div>
|
||
|
|
<div class="ttc" id="default__epilogue__tensor__op_8h_html"><div class="ttname"><a href="default__epilogue__tensor__op_8h.html">default_epilogue_tensor_op.h</a></div><div class="ttdoc">Epilogue for threadblock scoped GEMMs using Tensor Ops. </div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00_01Edd80343e6570718ed237122e4ebf7fb5_html_a9313414846dc9ae891e73793e4781c5c"><div class="ttname"><a href="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00_01Edd80343e6570718ed237122e4ebf7fb5.html#a9313414846dc9ae891e73793e4781c5c">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 >::Epilogue</a></div><div class="ttdeci">typename cutlass::epilogue::threadblock::DefaultEpilogueSimt< ThreadblockShape, typename Mma::Operator, EpilogueOutputOp, kEpilogueElementsPerAccess >::Epilogue Epilogue</div><div class="ttdoc">Define the epilogue. </div><div class="ttdef"><b>Definition:</b> default_gemm.h:434</div></div>
|
||
|
|
<div class="ttc" id="default__mma__core__sm75_8h_html"><div class="ttname"><a href="default__mma__core__sm75_8h.html">default_mma_core_sm75.h</a></div><div class="ttdoc">Defines basic properties needed by CTA-level GEMMs assuming expectations about data layout of the glo...</div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1epilogue_1_1threadblock_1_1DefaultEpilogueVoltaTensorOp_html"><div class="ttname"><a href="structcutlass_1_1epilogue_1_1threadblock_1_1DefaultEpilogueVoltaTensorOp.html">cutlass::epilogue::threadblock::DefaultEpilogueVoltaTensorOp</a></div><div class="ttdoc">Defines sensible defaults for epilogues for TensorOps. </div><div class="ttdef"><b>Definition:</b> default_epilogue_volta_tensor_op.h:71</div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00_01E5d78d37a9ae2ec08d7d477d571df036e_html_ac17af6195623850c325a2807e09f6b16"><div class="ttname"><a href="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00_01E5d78d37a9ae2ec08d7d477d571df036e.html#ac17af6195623850c325a2807e09f6b16">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 >::Epilogue</a></div><div class="ttdeci">typename cutlass::epilogue::threadblock::DefaultEpilogueTensorOp< ThreadblockShape, typename Mma::Operator, kPartitionsK, EpilogueOutputOp, EpilogueOutputOp::kCount >::Epilogue Epilogue</div><div class="ttdoc">Define the epilogue. </div><div class="ttdef"><b>Definition:</b> default_gemm.h:197</div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01ElementA_00_01layout_1_1ColumnMajorInterleave661fe54d13cc2c9153dcdf31e4beaa30_html_a783138201801f2fa054aa7bc21a8933b"><div class="ttname"><a href="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01ElementA_00_01layout_1_1ColumnMajorInterleave661fe54d13cc2c9153dcdf31e4beaa30.html#a783138201801f2fa054aa7bc21a8933b">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 >::Epilogue</a></div><div class="ttdeci">typename cutlass::epilogue::threadblock::DefaultInterleavedEpilogueTensorOp< ThreadblockShape, typename Mma::Operator, kPartitionsK, EpilogueOutputOp, 64/sizeof_bits< ElementC >::value, InterleavedK, IsBetaZero >::Epilogue Epilogue</div><div class="ttdoc">Define the epilogue. </div><div class="ttdef"><b>Definition:</b> default_gemm.h:261</div></div>
|
||
|
|
<div class="ttc" id="classcutlass_1_1layout_1_1RowMajor_html"><div class="ttname"><a href="classcutlass_1_1layout_1_1RowMajor.html">cutlass::layout::RowMajor</a></div><div class="ttdoc">Mapping function for row-major matrices. </div><div class="ttdef"><b>Definition:</b> layout/matrix.h:50</div></div>
|
||
|
|
<div class="ttc" id="epilogue_8h_html"><div class="ttname"><a href="epilogue_8h.html">epilogue.h</a></div><div class="ttdoc">Epilogue for threadblock scoped GEMMs using Tensor Ops. </div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1kernel_1_1Gemm_html"><div class="ttname"><a href="structcutlass_1_1gemm_1_1kernel_1_1Gemm.html">cutlass::gemm::kernel::Gemm</a></div><div class="ttdef"><b>Definition:</b> include/cutlass/gemm/kernel/gemm.h:52</div></div>
|
||
|
|
<div class="ttc" id="layout_2matrix_8h_html"><div class="ttname"><a href="layout_2matrix_8h.html">matrix.h</a></div><div class="ttdoc">Defines layout functions used by TensorRef and derived classes. </div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1epilogue_1_1threadblock_1_1DefaultEpilogueWmmaTensorOp_html"><div class="ttname"><a href="structcutlass_1_1epilogue_1_1threadblock_1_1DefaultEpilogueWmmaTensorOp.html">cutlass::epilogue::threadblock::DefaultEpilogueWmmaTensorOp</a></div><div class="ttdoc">Defines sensible defaults for epilogues for WMMA TensorOps. </div><div class="ttdef"><b>Definition:</b> default_epilogue_wmma_tensor_op.h:71</div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1epilogue_1_1threadblock_1_1DefaultInterleavedEpilogueTensorOp_html"><div class="ttname"><a href="structcutlass_1_1epilogue_1_1threadblock_1_1DefaultInterleavedEpilogueTensorOp.html">cutlass::epilogue::threadblock::DefaultInterleavedEpilogueTensorOp</a></div><div class="ttdef"><b>Definition:</b> default_epilogue_tensor_op.h:147</div></div>
|
||
|
|
<div class="ttc" id="gemm_2threadblock_2threadblock__swizzle_8h_html"><div class="ttname"><a href="gemm_2threadblock_2threadblock__swizzle_8h.html">threadblock_swizzle.h</a></div><div class="ttdoc">Implements several possible threadblock-swizzling functions mapping blockIdx to GEMM problems...</div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1layout_1_1ColumnMajorInterleaved_html"><div class="ttname"><a href="structcutlass_1_1layout_1_1ColumnMajorInterleaved.html">cutlass::layout::ColumnMajorInterleaved</a></div><div class="ttdef"><b>Definition:</b> layout/matrix.h:343</div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00_01Edd80343e6570718ed237122e4ebf7fb5_html_a2582e9547846a602ce69e9b9ba445393"><div class="ttname"><a href="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00_01Edd80343e6570718ed237122e4ebf7fb5.html#a2582e9547846a602ce69e9b9ba445393">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 >::Mma</a></div><div class="ttdeci">typename cutlass::gemm::threadblock::DefaultMma< ElementA, LayoutA, kAlignmentA, ElementB, LayoutB, kAlignmentB, ElementAccumulator, layout::RowMajor, arch::OpClassSimt, arch::Sm50, ThreadblockShape, WarpShape, GemmShape< 1, 1, 1 >, 2, Operator >::ThreadblockMma Mma</div><div class="ttdoc">Define the threadblock-scoped matrix multiply-accumulate. </div><div class="ttdef"><b>Definition:</b> default_gemm.h:423</div></div>
|
||
|
|
<div class="ttc" id="default__mma__core__simt_8h_html"><div class="ttname"><a href="default__mma__core__simt_8h.html">default_mma_core_simt.h</a></div><div class="ttdoc">Defines basic properties needed by CTA-level GEMMs assuming expectations about data layout of the glo...</div></div>
|
||
|
|
<div class="ttc" id="transform_2threadblock_2predicated__tile__iterator_8h_html"><div class="ttname"><a href="transform_2threadblock_2predicated__tile__iterator_8h.html">predicated_tile_iterator.h</a></div><div class="ttdoc">Templates implementing loading of tiles from pitch-linear rank=2 tensors. </div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_01inf48440732c1c5f42ddbfaba179861815_html_a1d734b7aae00825f298d3eb6e26adfbc"><div class="ttname"><a href="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_01inf48440732c1c5f42ddbfaba179861815.html#a1d734b7aae00825f298d3eb6e26adfbc">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 >::ElementA</a></div><div class="ttdeci">int8_t ElementA</div><div class="ttdef"><b>Definition:</b> default_gemm.h:481</div></div>
|
||
|
|
<div class="ttc" id="gemm__pipelined_8h_html"><div class="ttname"><a href="gemm__pipelined_8h.html">gemm_pipelined.h</a></div><div class="ttdoc">Template for a pipelined GEMM kernel. Does not compute batching or support split-K. </div></div>
|
||
|
|
<div class="ttc" id="wmma_8h_html"><div class="ttname"><a href="wmma_8h.html">wmma.h</a></div><div class="ttdoc">Templates exposing architecture support for warp matrix multiply-add (WMMA) operations. </div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1epilogue_1_1threadblock_1_1DefaultEpilogueTensorOp_html"><div class="ttname"><a href="structcutlass_1_1epilogue_1_1threadblock_1_1DefaultEpilogueTensorOp.html">cutlass::epilogue::threadblock::DefaultEpilogueTensorOp</a></div><div class="ttdoc">Defines sensible defaults for epilogues for TensorOps. </div><div class="ttdef"><b>Definition:</b> default_epilogue_tensor_op.h:72</div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_01inf48440732c1c5f42ddbfaba179861815_html_ae0d8ea41b9464dd460c7dc205fcaf1a3"><div class="ttname"><a href="structcutlass_1_1gemm_1_1kernel_1_1DefaultGemm_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_01inf48440732c1c5f42ddbfaba179861815.html#ae0d8ea41b9464dd460c7dc205fcaf1a3">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 >::Epilogue</a></div><div class="ttdeci">typename cutlass::epilogue::threadblock::DefaultEpilogueSimt< ThreadblockShape, typename Mma::Operator, EpilogueOutputOp, kEpilogueElementsPerAccess >::Epilogue Epilogue</div><div class="ttdoc">Define the epilogue. </div><div class="ttdef"><b>Definition:</b> default_gemm.h:513</div></div>
|
||
|
|
<div class="ttc" id="cutlass_8h_html"><div class="ttname"><a href="cutlass_8h.html">cutlass.h</a></div><div class="ttdoc">Basic include for CUTLASS. </div></div>
|
||
|
|
<div class="ttc" id="include_2cutlass_2gemm_2kernel_2gemm_8h_html"><div class="ttname"><a href="include_2cutlass_2gemm_2kernel_2gemm_8h.html">gemm.h</a></div><div class="ttdoc">Template for a pipelined GEMM kernel. Does not compute batching or support split-K. </div></div>
|
||
|
|
<div class="ttc" id="default__epilogue__simt_8h_html"><div class="ttname"><a href="default__epilogue__simt_8h.html">default_epilogue_simt.h</a></div><div class="ttdoc">Epilogue for threadblock scoped GEMMs using SIMT. </div></div>
|
||
|
|
<div class="ttc" id="default__epilogue__volta__tensor__op_8h_html"><div class="ttname"><a href="default__epilogue__volta__tensor__op_8h.html">default_epilogue_volta_tensor_op.h</a></div><div class="ttdoc">Epilogue for threadblock scoped GEMMs using Tensor Ops on Volta. </div></div>
|
||
|
|
<div class="ttc" id="structcutlass_1_1layout_1_1RowMajorInterleaved_html"><div class="ttname"><a href="structcutlass_1_1layout_1_1RowMajorInterleaved.html">cutlass::layout::RowMajorInterleaved</a></div><div class="ttdef"><b>Definition:</b> layout/matrix.h:237</div></div>
|
||
|
|
</div><!-- fragment --></div><!-- contents -->
|
||
|
|
<!-- start footer part -->
|
||
|
|
<hr class="footer"/><address class="footer"><small>
|
||
|
|
Generated by  <a href="http://www.doxygen.org/index.html">
|
||
|
|
<img class="footer" src="doxygen.png" alt="doxygen"/>
|
||
|
|
</a> 1.8.11
|
||
|
|
</small></address>
|
||
|
|
</body>
|
||
|
|
</html>
|