cutlass/docs/default__mma_8h_source.html

138 lines
80 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_mma.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&#160;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&#160;List</span></a></li>
<li><a href="globals.html"><span>File&#160;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_ffb18c781d484e5d1c680f712f01a439.html">threadblock</a></li> </ul>
</div>
</div><!-- top -->
<div class="header">
<div class="headertitle">
<div class="title">default_mma.h</div> </div>
</div><!--header-->
<div class="contents">
<a href="default__mma_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>&#160;<span class="comment">/***************************************************************************************************</span></div><div class="line"><a name="l00002"></a><span class="lineno"> 2</span>&#160;<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>&#160;<span class="comment"> *</span></div><div class="line"><a name="l00004"></a><span class="lineno"> 4</span>&#160;<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>&#160;<span class="comment"> * provided that the following conditions are met:</span></div><div class="line"><a name="l00006"></a><span class="lineno"> 6</span>&#160;<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>&#160;<span class="comment"> * conditions and the following disclaimer.</span></div><div class="line"><a name="l00008"></a><span class="lineno"> 8</span>&#160;<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>&#160;<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>&#160;<span class="comment"> * provided with the distribution.</span></div><div class="line"><a name="l00011"></a><span class="lineno"> 11</span>&#160;<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>&#160;<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>&#160;<span class="comment"> * permission.</span></div><div class="line"><a name="l00014"></a><span class="lineno"> 14</span>&#160;<span class="comment"> *</span></div><div class="line"><a name="l00015"></a><span class="lineno"> 15</span>&#160;<span class="comment"> * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS &quot;AS IS&quot; AND ANY EXPRESS OR</span></div><div class="line"><a name="l00016"></a><span class="lineno"> 16</span>&#160;<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>&#160;<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>&#160;<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>&#160;<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>&#160;<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>&#160;<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"></a><span class="lineno"> 22</span>&#160;<span class="comment"> * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.</span></div><div class="line"><a name="l00023"></a><span class="lineno"> 23</span>&#160;<span class="comment"> *</span></div><div class="line"><a name="l00024"></a><span class="lineno"> 24</span>&#160;<span class="comment"> **************************************************************************************************/</span></div><div class="line"><a name="l00029"></a><span class="lineno"> 29</span>&#160;<span class="preprocessor">#pragma once</span></div><div class="line"><a name="l00030"></a><span class="lineno"> 30</span>&#160;</div><div class="line"><a name="l00031"></a><span class="lineno"> 31</span>&#160;<span class="preprocessor">#include &quot;<a class="code" href="cutlass_8h.html">cutlass/cutlass.h</a>&quot;</span></div><div class="line"><a name="l00032"></a><span class="lineno"> 32</span>&#160;<span class="preprocessor">#include &quot;<a class="code" href="numeric__types_8h.html">cutlass/numeric_types.h</a>&quot;</span></div><div class="line"><a name="l00033"></a><span class="lineno"> 33</span>&#160;<span class="preprocessor">#include &quot;<a class="code" href="arch_8h.html">cutlass/arch/arch.h</a>&quot;</span></div><div class="line"><a name="l00034"></a><span class="lineno"> 34</span>&#160;<span class="preprocessor">#include &quot;<a class="code" href="wmma_8h.html">cutlass/arch/wmma.h</a>&quot;</span></div><div class="line"><a name="l00035"></a><span class="lineno"> 35</span>&#160;</div><div class="line"><a name="l00036"></a><span class="lineno"> 36</span>&#160;<span class="preprocessor">#include &quot;<a class="code" href="transform_2threadblock_2predicated__tile__iterator_8h.html">cutlass/transform/threadblock/predicated_tile_iterator.h</a>&quot;</span></div><div class="line"><a name="l00037"></a><span class="lineno"> 37</span>&#160;<span class="preprocessor">#include &quot;<a class="code" href="predicated__tile__iterator__2dthreadtile_8h.html">cutlass/transform/threadblock/predicated_tile_iterator_2dthreadtile.h</a>&quot;</span></div><div class="line"><a name="l00038"></a><span class="lineno"> 38</span>&#160;<span class="preprocessor">#include &quot;<a class="code" href="default__mma__core__sm70_8h.html">cutlass/gemm/threadblock/default_mma_core_sm70.h</a>&quot;</span></div><div class="line"><a name="l00039"></a><span class="lineno"> 39</span>&#160;<span class="preprocessor">#include &quot;<a class="code" href="default__mma__core__sm75_8h.html">cutlass/gemm/threadblock/default_mma_core_sm75.h</a>&quot;</span></div><div class="line"><a name="l00040"></a><span class="lineno"> 40</span>&#160;<span class="preprocessor">#if defined(CUTLASS_ARCH_WMMA_ENABLED)</span></div><div class="line"><a name="l00041"></a><span class="lineno"> 41</span>&#160;<span class="preprocessor">#include &quot;<a class="code" href="default__mma__core__wmma_8h.html">cutlass/gemm/threadblock/default_mma_core_wmma.h</a>&quot;</span></div><div class="line"><a name="l00042"></a><span class="lineno"> 42</span>&#160;<span class="preprocessor">#endif //CUTLASS_ARCH_WMMA_ENABLED</span></div><div class="line"><a name="l00043"></a><span class="lineno"> 43</span>&#160;</div><div class="line"><a name="l00045"></a><span class="lineno"> 45</span>&#160;</div><div class="line"><a name="l00046"></a><span class="lineno"> 46</span>&#160;<span class="keyword">namespace </span><a class="code" href="namespacecutlass.html">cutlass</a> {</div><div class="line"><a name="l00047"></a><span class="lineno"> 47</span>&#160;<span class="keyword">namespace </span>gemm {</div><div class="line"><a name="l00048"></a><span class="lineno"> 48</span>&#160;<span class="keyword">namespace </span>threadblock {</div><div class="line"><a name="l00049"></a><span class="lineno"> 49</span>&#160;</div><div class="line"><a name="l00051"></a><span class="lineno"> 51</span>&#160;</div><div class="line"><a name="l00052"></a><span class="lineno"> 52</span>&#160;<span class="keyword">template</span> &lt;</div><div class="line"><a name="l00054"></a><span class="lineno"> 54</span>&#160; <span class="keyword">typename</span> ElementA_,</div><div class="line"><a name="l00056"></a><span class="lineno"> 56</span>&#160; <span class="keyword">typename</span> LayoutA_,</div><div class="line"><a name="l00058"></a><span class="lineno"> 58</span>&#160; <span class="keywordtype">int</span> kAlignmentA,</div><div class="line"><a name="l00060"></a><span class="lineno"> 60</span>&#160; <span class="keyword">typename</span> ElementB_,</div><div class="line"><a name="l00062"></a><span class="lineno"> 62</span>&#160; <span class="keyword">typename</span> LayoutB_,</div><div class="line"><a name="l00064"></a><span class="lineno"> 64</span>&#160; <span class="keywordtype">int</span> kAlignmentB,</div><div class="line"><a name="l00066"></a><span class="lineno"> 66</span>&#160; <span class="keyword">typename</span> ElementAccumulator_,</div><div class="line"><a name="l00068"></a><span class="lineno"> 68</span>&#160; <span class="keyword">typename</span> LayoutC_,</div><div class="line"><a name="l00070"></a><span class="lineno"> 70</span>&#160; <span class="keyword">typename</span> OperatorClass_,</div><div class="line"><a name="l00072"></a><span class="lineno"> 72</span>&#160; <span class="keyword">typename</span> ArchTag_,</div><div class="line"><a name="l00074"></a><span class="lineno"> 74</span>&#160; <span class="keyword">typename</span> ThreadblockShape_,</div><div class="line"><a name="l00076"></a><span class="lineno"> 76</span>&#160; <span class="keyword">typename</span> WarpShape_,</div><div class="line"><a name="l00078"></a><span class="lineno"> 78</span>&#160; <span class="keyword">typename</span> InstructionShape_,</div><div class="line"><a name="l00080"></a><span class="lineno"> 80</span>&#160; <span class="keywordtype">int</span> Stages,</div><div class="line"><a name="l00082"></a><span class="lineno"> 82</span>&#160; <span class="keyword">typename</span> Operator,</div><div class="line"><a name="l00085"></a><span class="lineno"> 85</span>&#160; <span class="keywordtype">bool</span> AccumulatorsInRowMajor = <span class="keyword">false</span></div><div class="line"><a name="l00086"></a><span class="lineno"> 86</span>&#160; &gt;</div><div class="line"><a name="l00087"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma.html"> 87</a></span>&#160;<span class="keyword">struct </span><a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma.html">DefaultMma</a>;</div><div class="line"><a name="l00088"></a><span class="lineno"> 88</span>&#160;</div><div class="line"><a name="l00090"></a><span class="lineno"> 90</span>&#160;</div><div class="line"><a name="l00092"></a><span class="lineno"> 92</span>&#160;<span class="keyword">template</span> &lt;</div><div class="line"><a name="l00094"></a><span class="lineno"> 94</span>&#160; <span class="keyword">typename</span> ElementA,</div><div class="line"><a name="l00096"></a><span class="lineno"> 96</span>&#160; <span class="keyword">typename</span> LayoutA,</div><div class="line"><a name="l00098"></a><span class="lineno"> 98</span>&#160; <span class="keywordtype">int</span> kAlignmentA,</div><div class="line"><a name="l00100"></a><span class="lineno"> 100</span>&#160; <span class="keyword">typename</span> ElementB,</div><div class="line"><a name="l00102"></a><span class="lineno"> 102</span>&#160; <span class="keyword">typename</span> LayoutB,</div><div class="line"><a name="l00104"></a><span class="lineno"> 104</span>&#160; <span class="keywordtype">int</span> kAlignmentB,</div><div class="line"><a name="l00106"></a><span class="lineno"> 106</span>&#160; <span class="keyword">typename</span> ElementAccumulator,</div><div class="line"><a name="l00108"></a><span class="lineno"> 108</span>&#160; <span class="keyword">typename</span> ArchTag,</div><div class="line"><a name="l00110"></a><span class="lineno"> 110</span>&#160; <span class="keyword">typename</span> ThreadblockShape,</div><div class="line"><a name="l00112"></a><span class="lineno"> 112</span>&#160; <span class="keyword">typename</span> WarpShape,</div><div class="line"><a name="l00114"></a><span class="lineno"> 114</span>&#160; <span class="keyword">typename</span> InstructionShape,</div><div class="line"><a name="l00116"></a><span class="lineno"> 116</span>&#160; <span class="keyword">typename</span> Operator&gt;</div><div class="line"><a name="l00117"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00c67c16f9881e4f2fda76d8ed83ebabd6.html"> 117</a></span>&#160;<span class="keyword">struct </span><a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma.html">DefaultMma</a>&lt;ElementA, LayoutA, kAlignmentA, ElementB, LayoutB,</div><div class="line"><a name="l00118"></a><span class="lineno"> 118</span>&#160; kAlignmentB, ElementAccumulator, layout::RowMajor,</div><div class="line"><a name="l00119"></a><span class="lineno"> 119</span>&#160; arch::OpClassSimt, ArchTag, ThreadblockShape, WarpShape,</div><div class="line"><a name="l00120"></a><span class="lineno"> 120</span>&#160; InstructionShape, 2, Operator, false&gt; {</div><div class="line"><a name="l00121"></a><span class="lineno"> 121</span>&#160; <span class="comment">// Define the MmaCore components</span></div><div class="line"><a name="l00122"></a><span class="lineno"> 122</span>&#160; <span class="keyword">using</span> <a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00c67c16f9881e4f2fda76d8ed83ebabd6.html#a234366375e417a362efcc6ace66d59b6">MmaCore</a> = <span class="keyword">typename</span> <a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMmaCore.html">cutlass::gemm::threadblock::DefaultMmaCore</a>&lt;</div><div class="line"><a name="l00123"></a><span class="lineno"> 123</span>&#160; ThreadblockShape, WarpShape, InstructionShape, ElementA, LayoutA,</div><div class="line"><a name="l00124"></a><span class="lineno"> 124</span>&#160; ElementB, LayoutB, ElementAccumulator, <a class="code" href="classcutlass_1_1layout_1_1RowMajor.html">layout::RowMajor</a>,</div><div class="line"><a name="l00125"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00c67c16f9881e4f2fda76d8ed83ebabd6.html#a234366375e417a362efcc6ace66d59b6"> 125</a></span>&#160; arch::OpClassSimt, 2, Operator&gt;;</div><div class="line"><a name="l00126"></a><span class="lineno"> 126</span>&#160;</div><div class="line"><a name="l00127"></a><span class="lineno"> 127</span>&#160; <span class="comment">// Define iterators over tiles from the A operand</span></div><div class="line"><a name="l00128"></a><span class="lineno"> 128</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">IteratorA</a> =</div><div class="line"><a name="l00129"></a><span class="lineno"> 129</span>&#160; <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">cutlass::transform::threadblock::PredicatedTileIterator</a>&lt;</div><div class="line"><a name="l00130"></a><span class="lineno"> 130</span>&#160; <a class="code" href="structcutlass_1_1MatrixShape.html">cutlass::MatrixShape&lt;MmaCore::Shape::kM, MmaCore::Shape::kK&gt;</a>,</div><div class="line"><a name="l00131"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00c67c16f9881e4f2fda76d8ed83ebabd6.html#aed25b7eeff13af9a122fbf088c655137"> 131</a></span>&#160; ElementA, LayoutA, 1, <span class="keyword">typename</span> MmaCore::IteratorThreadMapA, kAlignmentA&gt;;</div><div class="line"><a name="l00132"></a><span class="lineno"> 132</span>&#160;</div><div class="line"><a name="l00133"></a><span class="lineno"> 133</span>&#160; <span class="comment">// Define iterators over tiles from the B operand</span></div><div class="line"><a name="l00134"></a><span class="lineno"> 134</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">IteratorB</a> =</div><div class="line"><a name="l00135"></a><span class="lineno"> 135</span>&#160; <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">cutlass::transform::threadblock::PredicatedTileIterator</a>&lt;</div><div class="line"><a name="l00136"></a><span class="lineno"> 136</span>&#160; <a class="code" href="structcutlass_1_1MatrixShape.html">cutlass::MatrixShape&lt;MmaCore::Shape::kK, MmaCore::Shape::kN&gt;</a>,</div><div class="line"><a name="l00137"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00c67c16f9881e4f2fda76d8ed83ebabd6.html#ad3cee865755ea98da79780c60a102f70"> 137</a></span>&#160; ElementB, LayoutB, 0, <span class="keyword">typename</span> MmaCore::IteratorThreadMapB, kAlignmentB&gt;;</div><div class="line"><a name="l00138"></a><span class="lineno"> 138</span>&#160;</div><div class="line"><a name="l00139"></a><span class="lineno"> 139</span>&#160; <span class="comment">// Define the threadblock-scoped pipelined matrix multiply</span></div><div class="line"><a name="l00140"></a><span class="lineno"> 140</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1gemm_1_1threadblock_1_1MmaPipelined.html">ThreadblockMma</a> = <a class="code" href="classcutlass_1_1gemm_1_1threadblock_1_1MmaPipelined.html">cutlass::gemm::threadblock::MmaPipelined</a>&lt;</div><div class="line"><a name="l00141"></a><span class="lineno"> 141</span>&#160; <span class="keyword">typename</span> MmaCore::Shape, <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">IteratorA</a>, <span class="keyword">typename</span> MmaCore::SmemIteratorA,</div><div class="line"><a name="l00142"></a><span class="lineno"> 142</span>&#160; <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">IteratorB</a>, <span class="keyword">typename</span> MmaCore::SmemIteratorB, ElementAccumulator,</div><div class="line"><a name="l00143"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00c67c16f9881e4f2fda76d8ed83ebabd6.html#a139c3d0243cd07b76d5e2a15b1f347aa"> 143</a></span>&#160; layout::RowMajor, <span class="keyword">typename</span> MmaCore::MmaPolicy&gt;;</div><div class="line"><a name="l00144"></a><span class="lineno"> 144</span>&#160;};</div><div class="line"><a name="l00145"></a><span class="lineno"> 145</span>&#160;</div><div class="line"><a name="l00146"></a><span class="lineno"> 146</span>&#160;</div><div class="line"><a name="l00148"></a><span class="lineno"> 148</span>&#160;<span class="keyword">template</span> &lt;</div><div class="line"><a name="l00150"></a><span class="lineno"> 150</span>&#160; <span class="keyword">typename</span> ElementA,</div><div class="line"><a name="l00152"></a><span class="lineno"> 152</span>&#160; <span class="keyword">typename</span> LayoutA,</div><div class="line"><a name="l00154"></a><span class="lineno"> 154</span>&#160; <span class="keywordtype">int</span> kAlignmentA,</div><div class="line"><a name="l00156"></a><span class="lineno"> 156</span>&#160; <span class="keyword">typename</span> ElementB,</div><div class="line"><a name="l00158"></a><span class="lineno"> 158</span>&#160; <span class="keyword">typename</span> LayoutB,</div><div class="line"><a name="l00160"></a><span class="lineno"> 160</span>&#160; <span class="keywordtype">int</span> kAlignmentB,</div><div class="line"><a name="l00162"></a><span class="lineno"> 162</span>&#160; <span class="keyword">typename</span> ElementAccumulator,</div><div class="line"><a name="l00164"></a><span class="lineno"> 164</span>&#160; <span class="keyword">typename</span> ArchTag,</div><div class="line"><a name="l00166"></a><span class="lineno"> 166</span>&#160; <span class="keyword">typename</span> ThreadblockShape,</div><div class="line"><a name="l00168"></a><span class="lineno"> 168</span>&#160; <span class="keyword">typename</span> WarpShape,</div><div class="line"><a name="l00170"></a><span class="lineno"> 170</span>&#160; <span class="keyword">typename</span> InstructionShape,</div><div class="line"><a name="l00172"></a><span class="lineno"> 172</span>&#160; <span class="keyword">typename</span> Operator</div><div class="line"><a name="l00173"></a><span class="lineno"> 173</span>&#160; &gt;</div><div class="line"><a name="l00174"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00ce36642cae579bce6605ff8edde3c6ab.html"> 174</a></span>&#160;<span class="keyword">struct </span><a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma.html">DefaultMma</a>&lt;ElementA, LayoutA, kAlignmentA, ElementB, LayoutB,</div><div class="line"><a name="l00175"></a><span class="lineno"> 175</span>&#160; kAlignmentB, ElementAccumulator, layout::RowMajor,</div><div class="line"><a name="l00176"></a><span class="lineno"> 176</span>&#160; arch::OpClassTensorOp, ArchTag, ThreadblockShape, WarpShape,</div><div class="line"><a name="l00177"></a><span class="lineno"> 177</span>&#160; InstructionShape, 2, Operator, false&gt; {</div><div class="line"><a name="l00178"></a><span class="lineno"> 178</span>&#160; <span class="comment">// Define the MmaCore components</span></div><div class="line"><a name="l00179"></a><span class="lineno"> 179</span>&#160; <span class="keyword">using</span> <a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00ce36642cae579bce6605ff8edde3c6ab.html#a84ba5d319587fe073f92e6e84a6dc9f7">MmaCore</a> = <span class="keyword">typename</span> <a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMmaCore.html">cutlass::gemm::threadblock::DefaultMmaCore</a>&lt;</div><div class="line"><a name="l00180"></a><span class="lineno"> 180</span>&#160; ThreadblockShape, WarpShape, InstructionShape, ElementA, LayoutA,</div><div class="line"><a name="l00181"></a><span class="lineno"> 181</span>&#160; ElementB, LayoutB, ElementAccumulator, <a class="code" href="classcutlass_1_1layout_1_1RowMajor.html">layout::RowMajor</a>,</div><div class="line"><a name="l00182"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00ce36642cae579bce6605ff8edde3c6ab.html#a84ba5d319587fe073f92e6e84a6dc9f7"> 182</a></span>&#160; arch::OpClassTensorOp, 2, Operator&gt;;</div><div class="line"><a name="l00183"></a><span class="lineno"> 183</span>&#160;</div><div class="line"><a name="l00184"></a><span class="lineno"> 184</span>&#160; <span class="comment">// Define iterators over tiles from the A operand</span></div><div class="line"><a name="l00185"></a><span class="lineno"> 185</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">IteratorA</a> =</div><div class="line"><a name="l00186"></a><span class="lineno"> 186</span>&#160; <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">cutlass::transform::threadblock::PredicatedTileIterator</a>&lt;</div><div class="line"><a name="l00187"></a><span class="lineno"> 187</span>&#160; <a class="code" href="structcutlass_1_1MatrixShape.html">cutlass::MatrixShape&lt;MmaCore::Shape::kM, MmaCore::Shape::kK&gt;</a>,</div><div class="line"><a name="l00188"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00ce36642cae579bce6605ff8edde3c6ab.html#a9c1ee7eaf97be7ac194306d11d8171e3"> 188</a></span>&#160; ElementA, LayoutA, 1, <span class="keyword">typename</span> MmaCore::IteratorThreadMapA, kAlignmentA&gt;;</div><div class="line"><a name="l00189"></a><span class="lineno"> 189</span>&#160;</div><div class="line"><a name="l00190"></a><span class="lineno"> 190</span>&#160; <span class="comment">// Define iterators over tiles from the B operand</span></div><div class="line"><a name="l00191"></a><span class="lineno"> 191</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">IteratorB</a> =</div><div class="line"><a name="l00192"></a><span class="lineno"> 192</span>&#160; <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">cutlass::transform::threadblock::PredicatedTileIterator</a>&lt;</div><div class="line"><a name="l00193"></a><span class="lineno"> 193</span>&#160; <a class="code" href="structcutlass_1_1MatrixShape.html">cutlass::MatrixShape&lt;MmaCore::Shape::kK, MmaCore::Shape::kN&gt;</a>,</div><div class="line"><a name="l00194"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00ce36642cae579bce6605ff8edde3c6ab.html#a845669c52313e80e4cd6c5f846332d64"> 194</a></span>&#160; ElementB, LayoutB, 0, <span class="keyword">typename</span> MmaCore::IteratorThreadMapB, kAlignmentB&gt;;</div><div class="line"><a name="l00195"></a><span class="lineno"> 195</span>&#160;</div><div class="line"><a name="l00196"></a><span class="lineno"> 196</span>&#160; <span class="comment">// Define the threadblock-scoped pipelined matrix multiply</span></div><div class="line"><a name="l00197"></a><span class="lineno"> 197</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1gemm_1_1threadblock_1_1MmaPipelined.html">ThreadblockMma</a> = <a class="code" href="classcutlass_1_1gemm_1_1threadblock_1_1MmaPipelined.html">cutlass::gemm::threadblock::MmaPipelined</a>&lt;</div><div class="line"><a name="l00198"></a><span class="lineno"> 198</span>&#160; <span class="keyword">typename</span> MmaCore::Shape, <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">IteratorA</a>, <span class="keyword">typename</span> MmaCore::SmemIteratorA,</div><div class="line"><a name="l00199"></a><span class="lineno"> 199</span>&#160; <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">IteratorB</a>, <span class="keyword">typename</span> MmaCore::SmemIteratorB, ElementAccumulator,</div><div class="line"><a name="l00200"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00ce36642cae579bce6605ff8edde3c6ab.html#abffacc8cdbb2092f32889a08d1a2ac63"> 200</a></span>&#160; layout::RowMajor, <span class="keyword">typename</span> MmaCore::MmaPolicy&gt;;</div><div class="line"><a name="l00201"></a><span class="lineno"> 201</span>&#160;};</div><div class="line"><a name="l00203"></a><span class="lineno"> 203</span>&#160;</div><div class="line"><a name="l00205"></a><span class="lineno"> 205</span>&#160;<span class="keyword">template</span> &lt;</div><div class="line"><a name="l00207"></a><span class="lineno"> 207</span>&#160; <span class="keyword">typename</span> ElementA,</div><div class="line"><a name="l00209"></a><span class="lineno"> 209</span>&#160; <span class="keyword">typename</span> LayoutA,</div><div class="line"><a name="l00211"></a><span class="lineno"> 211</span>&#160; <span class="keywordtype">int</span> kAlignmentA,</div><div class="line"><a name="l00213"></a><span class="lineno"> 213</span>&#160; <span class="keyword">typename</span> ElementB,</div><div class="line"><a name="l00215"></a><span class="lineno"> 215</span>&#160; <span class="keyword">typename</span> LayoutB,</div><div class="line"><a name="l00217"></a><span class="lineno"> 217</span>&#160; <span class="keywordtype">int</span> kAlignmentB,</div><div class="line"><a name="l00219"></a><span class="lineno"> 219</span>&#160; <span class="keyword">typename</span> ElementAccumulator,</div><div class="line"><a name="l00221"></a><span class="lineno"> 221</span>&#160; <span class="keyword">typename</span> OperatorClass,</div><div class="line"><a name="l00223"></a><span class="lineno"> 223</span>&#160; <span class="keyword">typename</span> ArchTag,</div><div class="line"><a name="l00225"></a><span class="lineno"> 225</span>&#160; <span class="keyword">typename</span> ThreadblockShape,</div><div class="line"><a name="l00227"></a><span class="lineno"> 227</span>&#160; <span class="keyword">typename</span> WarpShape,</div><div class="line"><a name="l00229"></a><span class="lineno"> 229</span>&#160; <span class="keyword">typename</span> InstructionShape,</div><div class="line"><a name="l00231"></a><span class="lineno"> 231</span>&#160; <span class="keyword">typename</span> Operator,</div><div class="line"><a name="l00233"></a><span class="lineno"> 233</span>&#160; <span class="keywordtype">int</span> InterleavedK&gt;</div><div class="line"><a name="l00234"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_0010764e1fd5a3251a57eddafbd83eab8e.html"> 234</a></span>&#160;<span class="keyword">struct </span><a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma.html">DefaultMma</a>&lt;ElementA, LayoutA, kAlignmentA, ElementB, LayoutB,</div><div class="line"><a name="l00235"></a><span class="lineno"> 235</span>&#160; kAlignmentB, ElementAccumulator,</div><div class="line"><a name="l00236"></a><span class="lineno"> 236</span>&#160; layout::ColumnMajorInterleaved&lt;InterleavedK&gt;, OperatorClass,</div><div class="line"><a name="l00237"></a><span class="lineno"> 237</span>&#160; ArchTag, ThreadblockShape, WarpShape, InstructionShape, 2,</div><div class="line"><a name="l00238"></a><span class="lineno"> 238</span>&#160; Operator, true&gt; {</div><div class="line"><a name="l00239"></a><span class="lineno"> 239</span>&#160; <span class="comment">// Define the MmaCore components</span></div><div class="line"><a name="l00240"></a><span class="lineno"> 240</span>&#160; <span class="keyword">using</span> <a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_0010764e1fd5a3251a57eddafbd83eab8e.html#afb9ea2c8962b31211307d744a6543451">MmaCore</a> = <span class="keyword">typename</span> <a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMmaCore.html">cutlass::gemm::threadblock::DefaultMmaCore</a>&lt;</div><div class="line"><a name="l00241"></a><span class="lineno"> 241</span>&#160; ThreadblockShape, WarpShape, InstructionShape, ElementA, LayoutA,</div><div class="line"><a name="l00242"></a><span class="lineno"> 242</span>&#160; ElementB, LayoutB, ElementAccumulator,</div><div class="line"><a name="l00243"></a><span class="lineno"> 243</span>&#160; <a class="code" href="structcutlass_1_1layout_1_1ColumnMajorInterleaved.html">layout::ColumnMajorInterleaved&lt;InterleavedK&gt;</a>, OperatorClass, 2, Operator,</div><div class="line"><a name="l00244"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_0010764e1fd5a3251a57eddafbd83eab8e.html#afb9ea2c8962b31211307d744a6543451"> 244</a></span>&#160; <span class="keyword">true</span>&gt;;</div><div class="line"><a name="l00245"></a><span class="lineno"> 245</span>&#160;</div><div class="line"><a name="l00246"></a><span class="lineno"> 246</span>&#160; <a class="code" href="platform_8h.html#adde4c9ea91b753491851361a4198c009">static_assert</a>(kAlignmentA == 128 / <a class="code" href="structcutlass_1_1sizeof__bits.html">sizeof_bits&lt;ElementA&gt;::value</a>, </div><div class="line"><a name="l00247"></a><span class="lineno"> 247</span>&#160; <span class="stringliteral">&quot;Alignment must match thread data map&#39;s vector length&quot;</span>);</div><div class="line"><a name="l00248"></a><span class="lineno"> 248</span>&#160;</div><div class="line"><a name="l00249"></a><span class="lineno"> 249</span>&#160; <a class="code" href="platform_8h.html#adde4c9ea91b753491851361a4198c009">static_assert</a>(kAlignmentB ==128 / <a class="code" href="structcutlass_1_1sizeof__bits.html">sizeof_bits&lt;ElementB&gt;::value</a>,</div><div class="line"><a name="l00250"></a><span class="lineno"> 250</span>&#160; <span class="stringliteral">&quot;Alignment must match thread data map&#39;s vector length&quot;</span>);</div><div class="line"><a name="l00251"></a><span class="lineno"> 251</span>&#160;</div><div class="line"><a name="l00252"></a><span class="lineno"> 252</span>&#160; <span class="comment">// Define iterators over tiles from the A operand</span></div><div class="line"><a name="l00253"></a><span class="lineno"> 253</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">IteratorA</a> = <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">cutlass::transform::threadblock::PredicatedTileIterator</a>&lt;</div><div class="line"><a name="l00254"></a><span class="lineno"> 254</span>&#160; <a class="code" href="structcutlass_1_1MatrixShape.html">cutlass::MatrixShape&lt;MmaCore::Shape::kM, MmaCore::Shape::kK&gt;</a>, ElementA,</div><div class="line"><a name="l00255"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_0010764e1fd5a3251a57eddafbd83eab8e.html#aa3a2bf4b00b6e5db0385c49f5a3136c8"> 255</a></span>&#160; LayoutA, 1, <span class="keyword">typename</span> MmaCore::IteratorThreadMapA&gt;;</div><div class="line"><a name="l00256"></a><span class="lineno"> 256</span>&#160;</div><div class="line"><a name="l00257"></a><span class="lineno"> 257</span>&#160; <span class="comment">// Define iterators over tiles from the B operand</span></div><div class="line"><a name="l00258"></a><span class="lineno"> 258</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">IteratorB</a> = <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">cutlass::transform::threadblock::PredicatedTileIterator</a>&lt;</div><div class="line"><a name="l00259"></a><span class="lineno"> 259</span>&#160; <a class="code" href="structcutlass_1_1MatrixShape.html">cutlass::MatrixShape&lt;MmaCore::Shape::kK, MmaCore::Shape::kN&gt;</a>, ElementB,</div><div class="line"><a name="l00260"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_0010764e1fd5a3251a57eddafbd83eab8e.html#a3c848ad770bb5309c2abbe63c19db604"> 260</a></span>&#160; LayoutB, 0, <span class="keyword">typename</span> MmaCore::IteratorThreadMapB&gt;;</div><div class="line"><a name="l00261"></a><span class="lineno"> 261</span>&#160;</div><div class="line"><a name="l00262"></a><span class="lineno"> 262</span>&#160; <span class="comment">// Define the threadblock-scoped pipelined matrix multiply</span></div><div class="line"><a name="l00263"></a><span class="lineno"> 263</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1gemm_1_1threadblock_1_1MmaPipelined.html">ThreadblockMma</a> = <a class="code" href="classcutlass_1_1gemm_1_1threadblock_1_1MmaPipelined.html">cutlass::gemm::threadblock::MmaPipelined</a>&lt;</div><div class="line"><a name="l00264"></a><span class="lineno"> 264</span>&#160; <span class="keyword">typename</span> MmaCore::Shape, <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">IteratorA</a>, <span class="keyword">typename</span> MmaCore::SmemIteratorA,</div><div class="line"><a name="l00265"></a><span class="lineno"> 265</span>&#160; <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">IteratorB</a>, <span class="keyword">typename</span> MmaCore::SmemIteratorB, ElementAccumulator,</div><div class="line"><a name="l00266"></a><span class="lineno"> 266</span>&#160; layout::ColumnMajorInterleaved&lt;InterleavedK&gt;,</div><div class="line"><a name="l00267"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_0010764e1fd5a3251a57eddafbd83eab8e.html#a0238bfafd32c806fe9f27c9d0b746f4d"> 267</a></span>&#160; <span class="keyword">typename</span> MmaCore::MmaPolicy&gt;;</div><div class="line"><a name="l00268"></a><span class="lineno"> 268</span>&#160;};</div><div class="line"><a name="l00269"></a><span class="lineno"> 269</span>&#160;</div><div class="line"><a name="l00273"></a><span class="lineno"> 273</span>&#160;<span class="keyword">template</span> &lt;</div><div class="line"><a name="l00275"></a><span class="lineno"> 275</span>&#160; <span class="keyword">typename</span> LayoutA,</div><div class="line"><a name="l00277"></a><span class="lineno"> 277</span>&#160; <span class="keywordtype">int</span> kAlignmentA,</div><div class="line"><a name="l00279"></a><span class="lineno"> 279</span>&#160; <span class="keyword">typename</span> LayoutB,</div><div class="line"><a name="l00281"></a><span class="lineno"> 281</span>&#160; <span class="keywordtype">int</span> kAlignmentB,</div><div class="line"><a name="l00283"></a><span class="lineno"> 283</span>&#160; <span class="keyword">typename</span> ElementAccumulator,</div><div class="line"><a name="l00285"></a><span class="lineno"> 285</span>&#160; <span class="keyword">typename</span> ArchTag,</div><div class="line"><a name="l00287"></a><span class="lineno"> 287</span>&#160; <span class="keyword">typename</span> ThreadblockShape,</div><div class="line"><a name="l00289"></a><span class="lineno"> 289</span>&#160; <span class="keyword">typename</span> Operator,</div><div class="line"><a name="l00291"></a><span class="lineno"> 291</span>&#160; <span class="keyword">typename</span> WarpShape&gt;</div><div class="line"><a name="l00292"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html"> 292</a></span>&#160;<span class="keyword">struct </span><a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma.html">DefaultMma</a>&lt;int8_t, LayoutA, kAlignmentA, int8_t, LayoutB, kAlignmentB,</div><div class="line"><a name="l00293"></a><span class="lineno"> 293</span>&#160; ElementAccumulator, layout::RowMajor, arch::OpClassSimt,</div><div class="line"><a name="l00294"></a><span class="lineno"> 294</span>&#160; ArchTag, ThreadblockShape, WarpShape, <a class="code" href="structcutlass_1_1gemm_1_1GemmShape.html">GemmShape</a>&lt;1, 1, 4&gt;, 2,</div><div class="line"><a name="l00295"></a><span class="lineno"> 295</span>&#160; Operator, false&gt; {</div><div class="line"><a name="l00296"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a465987b35751253ff48923ca24fc9457"> 296</a></span>&#160; <span class="keyword">using</span> InstructionShape = <a class="code" href="structcutlass_1_1gemm_1_1GemmShape.html">GemmShape&lt;1, 1, 4&gt;</a>;</div><div class="line"><a name="l00297"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a948ef845010e3f3c7e06a59ff8ad7439"> 297</a></span>&#160; <span class="keyword">using</span> <a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a948ef845010e3f3c7e06a59ff8ad7439">ElementA</a> = int8_t;</div><div class="line"><a name="l00298"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a3cac23bdfbef62d0964d3da236f39a36"> 298</a></span>&#160; <span class="keyword">using</span> ElementB = int8_t;</div><div class="line"><a name="l00299"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a0e76b54fea308a0820db78dcd3533d82"> 299</a></span>&#160; <span class="keyword">using</span> OperatorClass = arch::OpClassSimt;</div><div class="line"><a name="l00300"></a><span class="lineno"> 300</span>&#160;</div><div class="line"><a name="l00301"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a07e4567e2a2192830cc503c7199948e3"> 301</a></span>&#160; <span class="keyword">static</span> <span class="keyword">const</span> <span class="keywordtype">bool</span> transposeA = <a class="code" href="structcutlass_1_1platform_1_1is__same.html">cutlass::platform::is_same&lt; LayoutA, layout::ColumnMajor &gt;::value</a>;</div><div class="line"><a name="l00302"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a5288b71af92c940b1a9f89b99206d712"> 302</a></span>&#160; <span class="keyword">static</span> <span class="keyword">const</span> <span class="keywordtype">bool</span> transposeB = <a class="code" href="structcutlass_1_1platform_1_1is__same.html">cutlass::platform::is_same&lt; LayoutB, layout::RowMajor &gt;::value</a>;</div><div class="line"><a name="l00303"></a><span class="lineno"> 303</span>&#160;</div><div class="line"><a name="l00304"></a><span class="lineno"> 304</span>&#160; <span class="comment">// Define the MmaCore components</span></div><div class="line"><a name="l00305"></a><span class="lineno"> 305</span>&#160; <span class="keyword">using</span> <a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a0badaa2f2872010545f4f18e158dd4b2">MmaCore</a> = <span class="keyword">typename</span> <a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMmaCore.html">cutlass::gemm::threadblock::DefaultMmaCore</a>&lt;</div><div class="line"><a name="l00306"></a><span class="lineno"> 306</span>&#160; ThreadblockShape, WarpShape, InstructionShape, <a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a948ef845010e3f3c7e06a59ff8ad7439">ElementA</a>, LayoutA,</div><div class="line"><a name="l00307"></a><span class="lineno"> 307</span>&#160; ElementB, LayoutB, ElementAccumulator, <a class="code" href="classcutlass_1_1layout_1_1RowMajor.html">layout::RowMajor</a>,</div><div class="line"><a name="l00308"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a0badaa2f2872010545f4f18e158dd4b2"> 308</a></span>&#160; OperatorClass, 2, Operator&gt;;</div><div class="line"><a name="l00309"></a><span class="lineno"> 309</span>&#160;</div><div class="line"><a name="l00310"></a><span class="lineno"> 310</span>&#160; <span class="comment">// Define iterators over tiles from the A operand</span></div><div class="line"><a name="l00311"></a><span class="lineno"> 311</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator2dThreadTile.html">IteratorA</a> =</div><div class="line"><a name="l00312"></a><span class="lineno"> 312</span>&#160; <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator2dThreadTile.html">cutlass::transform::threadblock::PredicatedTileIterator2dThreadTile</a>&lt;</div><div class="line"><a name="l00313"></a><span class="lineno"> 313</span>&#160; <a class="code" href="structcutlass_1_1MatrixShape.html">cutlass::MatrixShape&lt;MmaCore::Shape::kM, MmaCore::Shape::kK&gt;</a>,</div><div class="line"><a name="l00314"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a42f485b11b46f1c9620a29c6095a865f"> 314</a></span>&#160; ElementA, LayoutA, 1, <span class="keyword">typename</span> MmaCore::IteratorThreadMapA, transposeA&gt;;</div><div class="line"><a name="l00315"></a><span class="lineno"> 315</span>&#160;</div><div class="line"><a name="l00316"></a><span class="lineno"> 316</span>&#160; <span class="comment">// Define iterators over tiles from the B operand</span></div><div class="line"><a name="l00317"></a><span class="lineno"> 317</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator2dThreadTile.html">IteratorB</a> =</div><div class="line"><a name="l00318"></a><span class="lineno"> 318</span>&#160; <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator2dThreadTile.html">cutlass::transform::threadblock::PredicatedTileIterator2dThreadTile</a>&lt;</div><div class="line"><a name="l00319"></a><span class="lineno"> 319</span>&#160; <a class="code" href="structcutlass_1_1MatrixShape.html">cutlass::MatrixShape&lt;MmaCore::Shape::kK, MmaCore::Shape::kN&gt;</a>,</div><div class="line"><a name="l00320"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#af5488e28ff5fec5af06f73b984be04b9"> 320</a></span>&#160; ElementB, LayoutB, 0, <span class="keyword">typename</span> MmaCore::IteratorThreadMapB, transposeB&gt;;</div><div class="line"><a name="l00321"></a><span class="lineno"> 321</span>&#160;</div><div class="line"><a name="l00322"></a><span class="lineno"> 322</span>&#160; <span class="comment">// Define the threadblock-scoped pipelined matrix multiply</span></div><div class="line"><a name="l00323"></a><span class="lineno"> 323</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1gemm_1_1threadblock_1_1MmaPipelined.html">ThreadblockMma</a> = <a class="code" href="classcutlass_1_1gemm_1_1threadblock_1_1MmaPipelined.html">cutlass::gemm::threadblock::MmaPipelined</a>&lt;</div><div class="line"><a name="l00324"></a><span class="lineno"> 324</span>&#160; <span class="keyword">typename</span> MmaCore::Shape, <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator2dThreadTile.html">IteratorA</a>, <span class="keyword">typename</span> MmaCore::SmemIteratorA,</div><div class="line"><a name="l00325"></a><span class="lineno"> 325</span>&#160; <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator2dThreadTile.html">IteratorB</a>, <span class="keyword">typename</span> MmaCore::SmemIteratorB, ElementAccumulator,</div><div class="line"><a name="l00326"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a284e8a5da0ed09ce4caf4da80f150849"> 326</a></span>&#160; layout::RowMajor, <span class="keyword">typename</span> MmaCore::MmaPolicy&gt;;</div><div class="line"><a name="l00327"></a><span class="lineno"> 327</span>&#160;};</div><div class="line"><a name="l00328"></a><span class="lineno"> 328</span>&#160;</div><div class="line"><a name="l00329"></a><span class="lineno"> 329</span>&#160;<span class="preprocessor">#if defined(CUTLASS_ARCH_WMMA_ENABLED)</span></div><div class="line"><a name="l00330"></a><span class="lineno"> 330</span>&#160;<span class="keyword">template</span> &lt;</div><div class="line"><a name="l00333"></a><span class="lineno"> 333</span>&#160; <span class="keyword">typename</span> <a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a948ef845010e3f3c7e06a59ff8ad7439">ElementA</a>,</div><div class="line"><a name="l00335"></a><span class="lineno"> 335</span>&#160; <span class="keyword">typename</span> LayoutA,</div><div class="line"><a name="l00337"></a><span class="lineno"> 337</span>&#160; <span class="keywordtype">int</span> kAlignmentA,</div><div class="line"><a name="l00339"></a><span class="lineno"> 339</span>&#160; <span class="keyword">typename</span> ElementB,</div><div class="line"><a name="l00341"></a><span class="lineno"> 341</span>&#160; <span class="keyword">typename</span> LayoutB,</div><div class="line"><a name="l00343"></a><span class="lineno"> 343</span>&#160; <span class="keywordtype">int</span> kAlignmentB,</div><div class="line"><a name="l00345"></a><span class="lineno"> 345</span>&#160; <span class="keyword">typename</span> ElementAccumulator,</div><div class="line"><a name="l00347"></a><span class="lineno"> 347</span>&#160; <span class="keyword">typename</span> LayoutC,</div><div class="line"><a name="l00349"></a><span class="lineno"> 349</span>&#160; <span class="keyword">typename</span> ArchTag,</div><div class="line"><a name="l00351"></a><span class="lineno"> 351</span>&#160; <span class="keyword">typename</span> ThreadblockShape,</div><div class="line"><a name="l00353"></a><span class="lineno"> 353</span>&#160; <span class="keyword">typename</span> WarpShape,</div><div class="line"><a name="l00355"></a><span class="lineno"> 355</span>&#160; <span class="keyword">typename</span> InstructionShape,</div><div class="line"><a name="l00357"></a><span class="lineno"> 357</span>&#160; <span class="keyword">typename</span> Operator&gt;</div><div class="line"><a name="l00358"></a><span class="lineno"> 358</span>&#160;<span class="keyword">struct </span><a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma.html">DefaultMma</a>&lt;<a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a948ef845010e3f3c7e06a59ff8ad7439">ElementA</a>, LayoutA, kAlignmentA, ElementB, LayoutB,</div><div class="line"><a name="l00359"></a><span class="lineno"> 359</span>&#160; kAlignmentB, ElementAccumulator, LayoutC,</div><div class="line"><a name="l00360"></a><span class="lineno"> 360</span>&#160; arch::OpClassWmmaTensorOp, ArchTag, ThreadblockShape, WarpShape,</div><div class="line"><a name="l00361"></a><span class="lineno"> 361</span>&#160; InstructionShape, 2, Operator&gt; {</div><div class="line"><a name="l00362"></a><span class="lineno"> 362</span>&#160; <span class="comment">// Define the MmaCore components</span></div><div class="line"><a name="l00363"></a><span class="lineno"> 363</span>&#160; <span class="keyword">using</span> <a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a0badaa2f2872010545f4f18e158dd4b2">MmaCore</a> = <span class="keyword">typename</span> <a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMmaCore.html">cutlass::gemm::threadblock::DefaultMmaCore</a>&lt;</div><div class="line"><a name="l00364"></a><span class="lineno"> 364</span>&#160; ThreadblockShape, WarpShape, InstructionShape, <a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a948ef845010e3f3c7e06a59ff8ad7439">ElementA</a>, LayoutA,</div><div class="line"><a name="l00365"></a><span class="lineno"> 365</span>&#160; ElementB, LayoutB, ElementAccumulator, LayoutC,</div><div class="line"><a name="l00366"></a><span class="lineno"> 366</span>&#160; arch::OpClassWmmaTensorOp, 2, Operator&gt;;</div><div class="line"><a name="l00367"></a><span class="lineno"> 367</span>&#160;</div><div class="line"><a name="l00368"></a><span class="lineno"> 368</span>&#160; <span class="comment">// Define iterators over tiles from the A operand</span></div><div class="line"><a name="l00369"></a><span class="lineno"> 369</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator2dThreadTile.html">IteratorA</a> =</div><div class="line"><a name="l00370"></a><span class="lineno"> 370</span>&#160; <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">cutlass::transform::threadblock::PredicatedTileIterator</a>&lt;</div><div class="line"><a name="l00371"></a><span class="lineno"> 371</span>&#160; <a class="code" href="structcutlass_1_1MatrixShape.html">cutlass::MatrixShape&lt;MmaCore::Shape::kM, MmaCore::Shape::kK&gt;</a>,</div><div class="line"><a name="l00372"></a><span class="lineno"> 372</span>&#160; ElementA, LayoutA, 1, <span class="keyword">typename</span> MmaCore::IteratorThreadMapA, kAlignmentA&gt;;</div><div class="line"><a name="l00373"></a><span class="lineno"> 373</span>&#160;</div><div class="line"><a name="l00374"></a><span class="lineno"> 374</span>&#160; <span class="comment">// Define iterators over tiles from the B operand</span></div><div class="line"><a name="l00375"></a><span class="lineno"> 375</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator2dThreadTile.html">IteratorB</a> =</div><div class="line"><a name="l00376"></a><span class="lineno"> 376</span>&#160; <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">cutlass::transform::threadblock::PredicatedTileIterator</a>&lt;</div><div class="line"><a name="l00377"></a><span class="lineno"> 377</span>&#160; <a class="code" href="structcutlass_1_1MatrixShape.html">cutlass::MatrixShape&lt;MmaCore::Shape::kK, MmaCore::Shape::kN&gt;</a>,</div><div class="line"><a name="l00378"></a><span class="lineno"> 378</span>&#160; ElementB, LayoutB, 0, <span class="keyword">typename</span> MmaCore::IteratorThreadMapB, kAlignmentB&gt;;</div><div class="line"><a name="l00379"></a><span class="lineno"> 379</span>&#160;</div><div class="line"><a name="l00380"></a><span class="lineno"> 380</span>&#160; <span class="comment">// Define the threadblock-scoped pipelined matrix multiply</span></div><div class="line"><a name="l00381"></a><span class="lineno"> 381</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1gemm_1_1threadblock_1_1MmaPipelined.html">ThreadblockMma</a> = <a class="code" href="classcutlass_1_1gemm_1_1threadblock_1_1MmaPipelined.html">cutlass::gemm::threadblock::MmaPipelined</a>&lt;</div><div class="line"><a name="l00382"></a><span class="lineno"> 382</span>&#160; <span class="keyword">typename</span> MmaCore::Shape, <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator2dThreadTile.html">IteratorA</a>, <span class="keyword">typename</span> MmaCore::SmemIteratorA,</div><div class="line"><a name="l00383"></a><span class="lineno"> 383</span>&#160; <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator2dThreadTile.html">IteratorB</a>, <span class="keyword">typename</span> MmaCore::SmemIteratorB, ElementAccumulator,</div><div class="line"><a name="l00384"></a><span class="lineno"> 384</span>&#160; LayoutC, <span class="keyword">typename</span> MmaCore::MmaPolicy&gt;;</div><div class="line"><a name="l00385"></a><span class="lineno"> 385</span>&#160;};</div><div class="line"><a name="l00386"></a><span class="lineno"> 386</span>&#160;</div><div class="line"><a name="l00388"></a><span class="lineno"> 388</span>&#160;<span class="keyword">template</span> &lt;</div><div class="line"><a name="l00390"></a><span class="lineno"> 390</span>&#160; <span class="keyword">typename</span> <a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a948ef845010e3f3c7e06a59ff8ad7439">ElementA</a>,</div><div class="line"><a name="l00392"></a><span class="lineno"> 392</span>&#160; <span class="keyword">typename</span> LayoutA,</div><div class="line"><a name="l00394"></a><span class="lineno"> 394</span>&#160; <span class="keywordtype">int</span> kAlignmentA,</div><div class="line"><a name="l00396"></a><span class="lineno"> 396</span>&#160; <span class="keyword">typename</span> ElementB,</div><div class="line"><a name="l00398"></a><span class="lineno"> 398</span>&#160; <span class="keyword">typename</span> LayoutB,</div><div class="line"><a name="l00400"></a><span class="lineno"> 400</span>&#160; <span class="keywordtype">int</span> kAlignmentB,</div><div class="line"><a name="l00402"></a><span class="lineno"> 402</span>&#160; <span class="keyword">typename</span> ElementAccumulator,</div><div class="line"><a name="l00404"></a><span class="lineno"> 404</span>&#160; <span class="keyword">typename</span> LayoutC,</div><div class="line"><a name="l00406"></a><span class="lineno"> 406</span>&#160; <span class="keyword">typename</span> ArchTag,</div><div class="line"><a name="l00408"></a><span class="lineno"> 408</span>&#160; <span class="keyword">typename</span> ThreadblockShape,</div><div class="line"><a name="l00410"></a><span class="lineno"> 410</span>&#160; <span class="keyword">typename</span> WarpShape,</div><div class="line"><a name="l00412"></a><span class="lineno"> 412</span>&#160; <span class="keyword">typename</span> InstructionShape,</div><div class="line"><a name="l00414"></a><span class="lineno"> 414</span>&#160; <span class="keyword">typename</span> Operator&gt;</div><div class="line"><a name="l00415"></a><span class="lineno"> 415</span>&#160;<span class="keyword">struct </span><a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma.html">DefaultMma</a>&lt;ElementA, LayoutA, kAlignmentA, ElementB, LayoutB,</div><div class="line"><a name="l00416"></a><span class="lineno"> 416</span>&#160; kAlignmentB, ElementAccumulator, LayoutC,</div><div class="line"><a name="l00417"></a><span class="lineno"> 417</span>&#160; arch::OpClassWmmaTensorOp, ArchTag, ThreadblockShape, WarpShape,</div><div class="line"><a name="l00418"></a><span class="lineno"> 418</span>&#160; InstructionShape, 1, Operator&gt; {</div><div class="line"><a name="l00419"></a><span class="lineno"> 419</span>&#160; <span class="comment">// Define the MmaCore components</span></div><div class="line"><a name="l00420"></a><span class="lineno"> 420</span>&#160; <span class="keyword">using</span> <a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a0badaa2f2872010545f4f18e158dd4b2">MmaCore</a> = <span class="keyword">typename</span> <a class="code" href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMmaCore.html">cutlass::gemm::threadblock::DefaultMmaCore</a>&lt;</div><div class="line"><a name="l00421"></a><span class="lineno"> 421</span>&#160; ThreadblockShape, WarpShape, InstructionShape, ElementA, LayoutA,</div><div class="line"><a name="l00422"></a><span class="lineno"> 422</span>&#160; ElementB, LayoutB, ElementAccumulator, LayoutC,</div><div class="line"><a name="l00423"></a><span class="lineno"> 423</span>&#160; arch::OpClassWmmaTensorOp, 1, Operator&gt;; </div><div class="line"><a name="l00424"></a><span class="lineno"> 424</span>&#160;</div><div class="line"><a name="l00425"></a><span class="lineno"> 425</span>&#160; <span class="comment">// Define iterators over tiles from the A operand</span></div><div class="line"><a name="l00426"></a><span class="lineno"> 426</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator2dThreadTile.html">IteratorA</a> =</div><div class="line"><a name="l00427"></a><span class="lineno"> 427</span>&#160; <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">cutlass::transform::threadblock::PredicatedTileIterator</a>&lt;</div><div class="line"><a name="l00428"></a><span class="lineno"> 428</span>&#160; <a class="code" href="structcutlass_1_1MatrixShape.html">cutlass::MatrixShape&lt;MmaCore::Shape::kM, MmaCore::Shape::kK&gt;</a>,</div><div class="line"><a name="l00429"></a><span class="lineno"> 429</span>&#160; ElementA, LayoutA, 1, <span class="keyword">typename</span> MmaCore::IteratorThreadMapA, kAlignmentA&gt;;</div><div class="line"><a name="l00430"></a><span class="lineno"> 430</span>&#160;</div><div class="line"><a name="l00431"></a><span class="lineno"> 431</span>&#160; <span class="comment">// Define iterators over tiles from the B operand</span></div><div class="line"><a name="l00432"></a><span class="lineno"> 432</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator2dThreadTile.html">IteratorB</a> =</div><div class="line"><a name="l00433"></a><span class="lineno"> 433</span>&#160; <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">cutlass::transform::threadblock::PredicatedTileIterator</a>&lt;</div><div class="line"><a name="l00434"></a><span class="lineno"> 434</span>&#160; <a class="code" href="structcutlass_1_1MatrixShape.html">cutlass::MatrixShape&lt;MmaCore::Shape::kK, MmaCore::Shape::kN&gt;</a>,</div><div class="line"><a name="l00435"></a><span class="lineno"> 435</span>&#160; ElementB, LayoutB, 0, <span class="keyword">typename</span> MmaCore::IteratorThreadMapB, kAlignmentB&gt;;</div><div class="line"><a name="l00436"></a><span class="lineno"> 436</span>&#160;</div><div class="line"><a name="l00437"></a><span class="lineno"> 437</span>&#160; <span class="comment">// Define the threadblock-scoped singlestage matrix multiply</span></div><div class="line"><a name="l00438"></a><span class="lineno"> 438</span>&#160; <span class="keyword">using</span> <a class="code" href="classcutlass_1_1gemm_1_1threadblock_1_1MmaPipelined.html">ThreadblockMma</a> = <a class="code" href="classcutlass_1_1gemm_1_1threadblock_1_1MmaSingleStage.html">cutlass::gemm::threadblock::MmaSingleStage</a>&lt;</div><div class="line"><a name="l00439"></a><span class="lineno"> 439</span>&#160; <span class="keyword">typename</span> MmaCore::Shape, <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator2dThreadTile.html">IteratorA</a>, <span class="keyword">typename</span> MmaCore::SmemIteratorA,</div><div class="line"><a name="l00440"></a><span class="lineno"> 440</span>&#160; <a class="code" href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator2dThreadTile.html">IteratorB</a>, <span class="keyword">typename</span> MmaCore::SmemIteratorB, ElementAccumulator,</div><div class="line"><a name="l00441"></a><span class="lineno"> 441</span>&#160; LayoutC, <span class="keyword">typename</span> MmaCore::MmaPolicy&gt;;</div><div class="line"><a name="l00442"></a><span class="lineno"> 442</span>&#160;};</div><div class="line"><a name="l00444"></a><span class="lineno"> 444</span>&#160;<span class="preprocessor">#endif //CUTLASS_ARCH_WMMA_ENABLED</span></div><div class="line"><a name="l00445"></a><span class="lineno"> 445</span>&#160;</div><div class="line"><a name="l00446"></a><span class="lineno"> 446</span>&#160;} <span class="comment">// namespace threadblock</span></div><div class="line"><a name="l00447"></a><span class="lineno"> 447</span>&#160;} <span class="comment">// namespace gemm</span></div><div class="line"><a name="l00448"></a><span class="lineno"> 448</span>&#160;} <span class="comment">// namespace cutlass </span></div><div class="line"><a name="l00449"></a><span class="lineno"> 449</span>&#160;</div><div class="ttc" id="structcutlass_1_1MatrixShape_html"><div class="ttname"><a href="structcutlass_1_1MatrixShape.html">cutlass::MatrixShape</a></div><div class="ttdoc">Describes the size of a matrix tile. </div><div class="ttdef"><b>Definition:</b> matrix_shape.h:42</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_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696_html_a0badaa2f2872010545f4f18e158dd4b2"><div class="ttname"><a href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a0badaa2f2872010545f4f18e158dd4b2">cutlass::gemm::threadblock::DefaultMma&lt; int8_t, LayoutA, kAlignmentA, int8_t, LayoutB, kAlignmentB, ElementAccumulator, layout::RowMajor, arch::OpClassSimt, ArchTag, ThreadblockShape, WarpShape, GemmShape&lt; 1, 1, 4 &gt;, 2, Operator, false &gt;::MmaCore</a></div><div class="ttdeci">typename cutlass::gemm::threadblock::DefaultMmaCore&lt; ThreadblockShape, WarpShape, InstructionShape, ElementA, LayoutA, ElementB, LayoutB, ElementAccumulator, layout::RowMajor, OperatorClass, 2, Operator &gt; MmaCore</div><div class="ttdef"><b>Definition:</b> default_mma.h:308</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1is__same_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1is__same.html">cutlass::platform::is_same</a></div><div class="ttdoc">std::is_same (false specialization) </div><div class="ttdef"><b>Definition:</b> platform.h:394</div></div>
<div class="ttc" id="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMmaCore_html"><div class="ttname"><a href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMmaCore.html">cutlass::gemm::threadblock::DefaultMmaCore</a></div><div class="ttdef"><b>Definition:</b> default_mma_core.h:90</div></div>
<div class="ttc" id="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696_html_a948ef845010e3f3c7e06a59ff8ad7439"><div class="ttname"><a href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01int8__t_00_01LayoutA_00_01kAlignmentA_00_07e7230d4011ada5e22cfcb29103b696.html#a948ef845010e3f3c7e06a59ff8ad7439">cutlass::gemm::threadblock::DefaultMma&lt; int8_t, LayoutA, kAlignmentA, int8_t, LayoutB, kAlignmentB, ElementAccumulator, layout::RowMajor, arch::OpClassSimt, ArchTag, ThreadblockShape, WarpShape, GemmShape&lt; 1, 1, 4 &gt;, 2, Operator, false &gt;::ElementA</a></div><div class="ttdeci">int8_t ElementA</div><div class="ttdef"><b>Definition:</b> default_mma.h:297</div></div>
<div class="ttc" id="default__mma__core__wmma_8h_html"><div class="ttname"><a href="default__mma__core__wmma_8h.html">default_mma_core_wmma.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="classcutlass_1_1gemm_1_1threadblock_1_1MmaPipelined_html"><div class="ttname"><a href="classcutlass_1_1gemm_1_1threadblock_1_1MmaPipelined.html">cutlass::gemm::threadblock::MmaPipelined</a></div><div class="ttdoc">Structure to compute the matrix product targeting CUDA cores and SIMT math instructions. </div><div class="ttdef"><b>Definition:</b> mma_pipelined.h:86</div></div>
<div class="ttc" id="classcutlass_1_1gemm_1_1threadblock_1_1MmaSingleStage_html"><div class="ttname"><a href="classcutlass_1_1gemm_1_1threadblock_1_1MmaSingleStage.html">cutlass::gemm::threadblock::MmaSingleStage</a></div><div class="ttdoc">Structure to compute the matrix product targeting CUDA cores and SIMT math instructions. </div><div class="ttdef"><b>Definition:</b> mma_singlestage.h:76</div></div>
<div class="ttc" id="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00c67c16f9881e4f2fda76d8ed83ebabd6_html_a234366375e417a362efcc6ace66d59b6"><div class="ttname"><a href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00c67c16f9881e4f2fda76d8ed83ebabd6.html#a234366375e417a362efcc6ace66d59b6">cutlass::gemm::threadblock::DefaultMma&lt; ElementA, LayoutA, kAlignmentA, ElementB, LayoutB, kAlignmentB, ElementAccumulator, layout::RowMajor, arch::OpClassSimt, ArchTag, ThreadblockShape, WarpShape, InstructionShape, 2, Operator, false &gt;::MmaCore</a></div><div class="ttdeci">typename cutlass::gemm::threadblock::DefaultMmaCore&lt; ThreadblockShape, WarpShape, InstructionShape, ElementA, LayoutA, ElementB, LayoutB, ElementAccumulator, layout::RowMajor, arch::OpClassSimt, 2, Operator &gt; MmaCore</div><div class="ttdef"><b>Definition:</b> default_mma.h:125</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_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_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00ce36642cae579bce6605ff8edde3c6ab_html_a84ba5d319587fe073f92e6e84a6dc9f7"><div class="ttname"><a href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_00ce36642cae579bce6605ff8edde3c6ab.html#a84ba5d319587fe073f92e6e84a6dc9f7">cutlass::gemm::threadblock::DefaultMma&lt; ElementA, LayoutA, kAlignmentA, ElementB, LayoutB, kAlignmentB, ElementAccumulator, layout::RowMajor, arch::OpClassTensorOp, ArchTag, ThreadblockShape, WarpShape, InstructionShape, 2, Operator, false &gt;::MmaCore</a></div><div class="ttdeci">typename cutlass::gemm::threadblock::DefaultMmaCore&lt; ThreadblockShape, WarpShape, InstructionShape, ElementA, LayoutA, ElementB, LayoutB, ElementAccumulator, layout::RowMajor, arch::OpClassTensorOp, 2, Operator &gt; MmaCore</div><div class="ttdef"><b>Definition:</b> default_mma.h:182</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__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="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator_html"><div class="ttname"><a href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator.html">cutlass::transform::threadblock::PredicatedTileIterator</a></div><div class="ttdef"><b>Definition:</b> transform/threadblock/predicated_tile_iterator.h:133</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="arch_8h_html"><div class="ttname"><a href="arch_8h.html">arch.h</a></div><div class="ttdoc">Defines tags for architecture-specific configurations. </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="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_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_0010764e1fd5a3251a57eddafbd83eab8e_html_afb9ea2c8962b31211307d744a6543451"><div class="ttname"><a href="structcutlass_1_1gemm_1_1threadblock_1_1DefaultMma_3_01ElementA_00_01LayoutA_00_01kAlignmentA_0010764e1fd5a3251a57eddafbd83eab8e.html#afb9ea2c8962b31211307d744a6543451">cutlass::gemm::threadblock::DefaultMma&lt; ElementA, LayoutA, kAlignmentA, ElementB, LayoutB, kAlignmentB, ElementAccumulator, layout::ColumnMajorInterleaved&lt; InterleavedK &gt;, OperatorClass, ArchTag, ThreadblockShape, WarpShape, InstructionShape, 2, Operator, true &gt;::MmaCore</a></div><div class="ttdeci">typename cutlass::gemm::threadblock::DefaultMmaCore&lt; ThreadblockShape, WarpShape, InstructionShape, ElementA, LayoutA, ElementB, LayoutB, ElementAccumulator, layout::ColumnMajorInterleaved&lt; InterleavedK &gt;, OperatorClass, 2, Operator, true &gt; MmaCore</div><div class="ttdef"><b>Definition:</b> default_mma.h:244</div></div>
<div class="ttc" id="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator2dThreadTile_html"><div class="ttname"><a href="classcutlass_1_1transform_1_1threadblock_1_1PredicatedTileIterator2dThreadTile.html">cutlass::transform::threadblock::PredicatedTileIterator2dThreadTile</a></div><div class="ttdef"><b>Definition:</b> predicated_tile_iterator_2dthreadtile.h:133</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="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="predicated__tile__iterator__2dthreadtile_8h_html"><div class="ttname"><a href="predicated__tile__iterator__2dthreadtile_8h.html">predicated_tile_iterator_2dthreadtile.h</a></div><div class="ttdoc">Templates implementing loading of tiles from pitch-linear rank=2 tensors. </div></div>
</div><!-- fragment --></div><!-- contents -->
<!-- start footer part -->
<hr class="footer"/><address class="footer"><small>
Generated by &#160;<a href="http://www.doxygen.org/index.html">
<img class="footer" src="doxygen.png" alt="doxygen"/>
</a> 1.8.11
</small></address>
</body>
</html>