109 lines
34 KiB
HTML
109 lines
34 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 1.8.14"/>
|
|
<meta name="viewport" content="width=device-width, initial-scale=1"/>
|
|
<title>Cutlass: hgemm_multiply_add.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/x-mathjax-config">
|
|
MathJax.Hub.Config({
|
|
extensions: ["tex2jax.js"],
|
|
jax: ["input/TeX","output/HTML-CSS"],
|
|
});
|
|
</script><script type="text/javascript" async 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="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 1.8.14 -->
|
|
<script type="text/javascript">
|
|
/* @license magnet:?xt=urn:btih:cf05388f2679ee054f2beb29a391d25f4e673ac3&dn=gpl-2.0.txt GPL-v2 */
|
|
var searchBox = new SearchBox("searchBox", "search",false,'Search');
|
|
/* @license-end */
|
|
</script>
|
|
<script type="text/javascript" src="menudata.js"></script>
|
|
<script type="text/javascript" src="menu.js"></script>
|
|
<script type="text/javascript">
|
|
/* @license magnet:?xt=urn:btih:cf05388f2679ee054f2beb29a391d25f4e673ac3&dn=gpl-2.0.txt GPL-v2 */
|
|
$(function() {
|
|
initMenu('',true,false,'search.php','Search');
|
|
$(document).ready(function() { init_search(); });
|
|
});
|
|
/* @license-end */</script>
|
|
<div id="main-nav"></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_1417ee5ebebc309c36b7962f26a92c39.html">cutlass</a></li><li class="navelem"><a class="el" href="dir_18d6a367a3982a494d65599933fc67a3.html">gemm</a></li> </ul>
|
|
</div>
|
|
</div><!-- top -->
|
|
<div class="header">
|
|
<div class="headertitle">
|
|
<div class="title">hgemm_multiply_add.h</div> </div>
|
|
</div><!--header-->
|
|
<div class="contents">
|
|
<a href="hgemm__multiply__add_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) 2017-2018, 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"></a><span class="lineno"> 22</span> <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> <span class="comment"> *</span></div><div class="line"><a name="l00024"></a><span class="lineno"> 24</span> <span class="comment"> **************************************************************************************************/</span></div><div class="line"><a name="l00029"></a><span class="lineno"> 29</span> <span class="preprocessor">#pragma once</span></div><div class="line"><a name="l00030"></a><span class="lineno"> 30</span> </div><div class="line"><a name="l00031"></a><span class="lineno"> 31</span> <span class="preprocessor">#include "<a class="code" href="fragment_8h.html">cutlass/fragment.h</a>"</span></div><div class="line"><a name="l00032"></a><span class="lineno"> 32</span> </div><div class="line"><a name="l00033"></a><span class="lineno"> 33</span> <span class="preprocessor">#include "<a class="code" href="thread__multiply__add_8h.html">cutlass/gemm/thread_multiply_add.h</a>"</span></div><div class="line"><a name="l00034"></a><span class="lineno"> 34</span> </div><div class="line"><a name="l00035"></a><span class="lineno"> 35</span> <span class="keyword">namespace </span><a class="code" href="namespacecutlass.html">cutlass</a> {</div><div class="line"><a name="l00036"></a><span class="lineno"> 36</span> <span class="keyword">namespace </span>gemm {</div><div class="line"><a name="l00037"></a><span class="lineno"> 37</span> </div><div class="line"><a name="l00039"></a><span class="lineno"> 39</span> </div><div class="line"><a name="l00041"></a><span class="lineno"> 41</span> <span class="keyword">template</span> <<span class="keyword">typename</span> ThreadGemmShape_, <span class="keyword">typename</span> ThreadsPerWarp_></div><div class="line"><a name="l00042"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html"> 42</a></span> <span class="keyword">struct </span><a class="code" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd.html">ThreadMultiplyAdd</a><ThreadGemmShape_, ThreadsPerWarp_, half, half, half> {</div><div class="line"><a name="l00044"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a245bdf059794144d04d2823d2b39588e"> 44</a></span>  <span class="keyword">typedef</span> <a class="code" href="structcutlass_1_1Shape.html">Shape<1, 1, 2, 1></a> <a class="code" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a245bdf059794144d04d2823d2b39588e">InstructionShape</a>;</div><div class="line"><a name="l00046"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#ac095b403212e23cb95e70cee9013099e"> 46</a></span>  <span class="keyword">typedef</span> ThreadGemmShape_ <a class="code" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#ac095b403212e23cb95e70cee9013099e">ThreadGemmShape</a>;</div><div class="line"><a name="l00048"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a24dd9fdc54b001840e8b82664b3bfe3a"> 48</a></span>  <span class="keyword">typedef</span> <a class="code" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#ac095b403212e23cb95e70cee9013099e">ThreadGemmShape</a> <a class="code" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a24dd9fdc54b001840e8b82664b3bfe3a">AccumulatorsPerThread</a>;</div><div class="line"><a name="l00050"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a61ed3d8bb4ca6db39d16b632f58d75dc"> 50</a></span>  <span class="keyword">typedef</span> ThreadsPerWarp_ <a class="code" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a61ed3d8bb4ca6db39d16b632f58d75dc">ThreadsPerWarp</a>;</div><div class="line"><a name="l00052"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#aacb9a5a1d8f3b6e21bc449b0b97949b7"> 52</a></span>  <span class="keyword">typedef</span> <span class="keyword">typename</span> <a class="code" href="structcutlass_1_1ShapeMul.html#a8875fc5e861339f981360ed774e8cc94">ShapeMul<ThreadGemmShape, ThreadsPerWarp>::Shape</a> <a class="code" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#aacb9a5a1d8f3b6e21bc449b0b97949b7">AccumulatorsPerWarp</a>;</div><div class="line"><a name="l00054"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a1a8f6feed85c7e88b36bc1a2637c716f"> 54</a></span>  <span class="keyword">typedef</span> half <a class="code" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a1a8f6feed85c7e88b36bc1a2637c716f">ScalarA</a>;</div><div class="line"><a name="l00056"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a1889bdc9e88265a8afdaeeca217a3372"> 56</a></span>  <span class="keyword">typedef</span> <a class="code" href="structcutlass_1_1Fragment.html">Fragment<ScalarA, AccumulatorsPerThread::kW></a> <a class="code" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a1889bdc9e88265a8afdaeeca217a3372">FragmentA</a>;</div><div class="line"><a name="l00058"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a70dfd2f33548dbd104d798f728526fbc"> 58</a></span>  <span class="keyword">typedef</span> half <a class="code" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a70dfd2f33548dbd104d798f728526fbc">ScalarB</a>;</div><div class="line"><a name="l00060"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#ae8f4db1465f5e082c6855bf13a4751c7"> 60</a></span>  <span class="keyword">typedef</span> <a class="code" href="structcutlass_1_1Fragment.html">Fragment<ScalarB, AccumulatorsPerThread::kH></a> <a class="code" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#ae8f4db1465f5e082c6855bf13a4751c7">FragmentB</a>;</div><div class="line"><a name="l00062"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a48a641d601c88d95aa542b636f94d60d"> 62</a></span>  <span class="keyword">typedef</span> half <a class="code" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a48a641d601c88d95aa542b636f94d60d">ScalarC</a>;</div><div class="line"><a name="l00064"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#af8124acb485709dba1c5378faa24516c"> 64</a></span>  <span class="keyword">typedef</span> <a class="code" href="structcutlass_1_1Fragment.html">Fragment<half, AccumulatorsPerThread::kH * AccumulatorsPerThread::kW></a> <a class="code" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#af8124acb485709dba1c5378faa24516c">Accumulators</a>;</div><div class="line"><a name="l00065"></a><span class="lineno"> 65</span> </div><div class="line"><a name="l00067"></a><span class="lineno"> 67</span>  <a class="code" href="platform_8h.html#adde4c9ea91b753491851361a4198c009">static_assert</a>(AccumulatorsPerThread::kH % 2 == 0, <span class="stringliteral">"Invalid size"</span>);</div><div class="line"><a name="l00068"></a><span class="lineno"> 68</span>  <a class="code" href="platform_8h.html#adde4c9ea91b753491851361a4198c009">static_assert</a>(AccumulatorsPerThread::kW % 2 == 0, <span class="stringliteral">"Invalid size"</span>);</div><div class="line"><a name="l00069"></a><span class="lineno"> 69</span> </div><div class="line"><a name="l00071"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a32b234c873ffe44090a12e12d871024c"> 71</a></span>  CUTLASS_DEVICE <a class="code" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a32b234c873ffe44090a12e12d871024c">ThreadMultiplyAdd</a>() {}</div><div class="line"><a name="l00072"></a><span class="lineno"> 72</span> </div><div class="line"><a name="l00074"></a><span class="lineno"><a class="line" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#aaf8ea4aaf393488f03ba7fb7af264940"> 74</a></span>  CUTLASS_DEVICE <span class="keywordtype">void</span> <a class="code" href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#aaf8ea4aaf393488f03ba7fb7af264940">multiply_add</a>(<a class="code" href="structcutlass_1_1Fragment.html">FragmentA</a> <span class="keyword">const</span>& a,</div><div class="line"><a name="l00075"></a><span class="lineno"> 75</span>  <a class="code" href="structcutlass_1_1Fragment.html">FragmentB</a> <span class="keyword">const</span>& b,</div><div class="line"><a name="l00076"></a><span class="lineno"> 76</span>  <a class="code" href="structcutlass_1_1Fragment.html">Accumulators</a> <span class="keyword">const</span>& c,</div><div class="line"><a name="l00077"></a><span class="lineno"> 77</span>  <a class="code" href="structcutlass_1_1Fragment.html">Accumulators</a>& d) {</div><div class="line"><a name="l00078"></a><span class="lineno"> 78</span> <span class="preprocessor">#if defined(__CUDACC__) && __CUDA_ARCH__ >= 530</span></div><div class="line"><a name="l00079"></a><span class="lineno"> 79</span>  <span class="comment">// The inputs.</span></div><div class="line"><a name="l00080"></a><span class="lineno"> 80</span>  __half2 <span class="keyword">const</span>* a_half2 = <span class="keyword">reinterpret_cast<</span>__half2 const*<span class="keyword">></span>(&a[0]);</div><div class="line"><a name="l00081"></a><span class="lineno"> 81</span>  __half2 <span class="keyword">const</span>* b_half2 = <span class="keyword">reinterpret_cast<</span>__half2 const*<span class="keyword">></span>(&b[0]);</div><div class="line"><a name="l00082"></a><span class="lineno"> 82</span>  __half2 <span class="keyword">const</span>* c_half2 = <span class="keyword">reinterpret_cast<</span>__half2 const*<span class="keyword">></span>(&c[0]);</div><div class="line"><a name="l00083"></a><span class="lineno"> 83</span> </div><div class="line"><a name="l00084"></a><span class="lineno"> 84</span>  <span class="comment">// The output.</span></div><div class="line"><a name="l00085"></a><span class="lineno"> 85</span>  __half2* d_half2 = <span class="keyword">reinterpret_cast<</span>__half2*<span class="keyword">></span>(&d[0]);</div><div class="line"><a name="l00086"></a><span class="lineno"> 86</span> </div><div class="line"><a name="l00087"></a><span class="lineno"> 87</span>  <span class="keywordflow">for</span> (<span class="keywordtype">int</span> j = 0; j < AccumulatorsPerThread::kH / 2; ++j) {</div><div class="line"><a name="l00088"></a><span class="lineno"> 88</span>  <span class="keywordflow">for</span> (<span class="keywordtype">int</span> i = 0; i < AccumulatorsPerThread::kW / 2; ++i) {</div><div class="line"><a name="l00089"></a><span class="lineno"> 89</span>  <span class="comment">// The offsets in the output fragment.</span></div><div class="line"><a name="l00090"></a><span class="lineno"> 90</span>  <span class="keywordtype">int</span> <span class="keyword">const</span> k0 = (2 * j + 0) * (AccumulatorsPerThread::kW / 2) + i;</div><div class="line"><a name="l00091"></a><span class="lineno"> 91</span>  <span class="keywordtype">int</span> <span class="keyword">const</span> k1 = (2 * j + 1) * (AccumulatorsPerThread::kW / 2) + i;</div><div class="line"><a name="l00092"></a><span class="lineno"> 92</span> </div><div class="line"><a name="l00093"></a><span class="lineno"> 93</span>  <span class="comment">// Compute the product a[i] * b[j].low.</span></div><div class="line"><a name="l00094"></a><span class="lineno"> 94</span>  d_half2[k0] = __hfma2(a_half2[i], __low2half2(b_half2[j]), c_half2[k0]);</div><div class="line"><a name="l00095"></a><span class="lineno"> 95</span>  <span class="comment">// Compute the product a[i] * b[j].high.</span></div><div class="line"><a name="l00096"></a><span class="lineno"> 96</span>  d_half2[k1] = __hfma2(a_half2[i], __high2half2(b_half2[j]), c_half2[k1]);</div><div class="line"><a name="l00097"></a><span class="lineno"> 97</span>  }</div><div class="line"><a name="l00098"></a><span class="lineno"> 98</span>  }</div><div class="line"><a name="l00099"></a><span class="lineno"> 99</span> <span class="preprocessor">#endif</span></div><div class="line"><a name="l00100"></a><span class="lineno"> 100</span>  }</div><div class="line"><a name="l00101"></a><span class="lineno"> 101</span> };</div><div class="line"><a name="l00102"></a><span class="lineno"> 102</span> </div><div class="line"><a name="l00104"></a><span class="lineno"> 104</span> </div><div class="line"><a name="l00105"></a><span class="lineno"> 105</span> } <span class="comment">// namespace gemm</span></div><div class="line"><a name="l00106"></a><span class="lineno"> 106</span> } <span class="comment">// namespace cutlass</span></div><div class="ttc" id="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4_html_a32b234c873ffe44090a12e12d871024c"><div class="ttname"><a href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a32b234c873ffe44090a12e12d871024c">cutlass::gemm::ThreadMultiplyAdd< ThreadGemmShape_, ThreadsPerWarp_, half, half, half >::ThreadMultiplyAdd</a></div><div class="ttdeci">CUTLASS_DEVICE ThreadMultiplyAdd()</div><div class="ttdoc">Make sure there&#39;s an even number of elements in both dimensions. </div><div class="ttdef"><b>Definition:</b> hgemm_multiply_add.h:71</div></div>
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4_html_a48a641d601c88d95aa542b636f94d60d"><div class="ttname"><a href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a48a641d601c88d95aa542b636f94d60d">cutlass::gemm::ThreadMultiplyAdd< ThreadGemmShape_, ThreadsPerWarp_, half, half, half >::ScalarC</a></div><div class="ttdeci">half ScalarC</div><div class="ttdoc">The type for C and D. </div><div class="ttdef"><b>Definition:</b> hgemm_multiply_add.h:62</div></div>
|
|
<div class="ttc" id="namespacecutlass_html"><div class="ttname"><a href="namespacecutlass.html">cutlass</a></div><div class="ttdef"><b>Definition:</b> convert.h:33</div></div>
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4_html_ae8f4db1465f5e082c6855bf13a4751c7"><div class="ttname"><a href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#ae8f4db1465f5e082c6855bf13a4751c7">cutlass::gemm::ThreadMultiplyAdd< ThreadGemmShape_, ThreadsPerWarp_, half, half, half >::FragmentB</a></div><div class="ttdeci">Fragment< ScalarB, AccumulatorsPerThread::kH > FragmentB</div><div class="ttdoc">The fragment for B. </div><div class="ttdef"><b>Definition:</b> hgemm_multiply_add.h:60</div></div>
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4_html_ac095b403212e23cb95e70cee9013099e"><div class="ttname"><a href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#ac095b403212e23cb95e70cee9013099e">cutlass::gemm::ThreadMultiplyAdd< ThreadGemmShape_, ThreadsPerWarp_, half, half, half >::ThreadGemmShape</a></div><div class="ttdeci">ThreadGemmShape_ ThreadGemmShape</div><div class="ttdoc">The number of accumulators per thread. </div><div class="ttdef"><b>Definition:</b> hgemm_multiply_add.h:46</div></div>
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4_html_a70dfd2f33548dbd104d798f728526fbc"><div class="ttname"><a href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a70dfd2f33548dbd104d798f728526fbc">cutlass::gemm::ThreadMultiplyAdd< ThreadGemmShape_, ThreadsPerWarp_, half, half, half >::ScalarB</a></div><div class="ttdeci">half ScalarB</div><div class="ttdoc">The type for B. </div><div class="ttdef"><b>Definition:</b> hgemm_multiply_add.h:58</div></div>
|
|
<div class="ttc" id="structcutlass_1_1ShapeMul_html_a8875fc5e861339f981360ed774e8cc94"><div class="ttname"><a href="structcutlass_1_1ShapeMul.html#a8875fc5e861339f981360ed774e8cc94">cutlass::ShapeMul::Shape</a></div><div class="ttdeci">Shape< A_::kD *B_::kD, A_::kH *B_::kH, A_::kW *B_::kW, A_::kC *B_::kC > Shape</div><div class="ttdef"><b>Definition:</b> shape.h:119</div></div>
|
|
<div class="ttc" id="structcutlass_1_1Fragment_html"><div class="ttname"><a href="structcutlass_1_1Fragment.html">cutlass::Fragment</a></div><div class="ttdoc">A template defining Fragment Concept. </div><div class="ttdef"><b>Definition:</b> fragment.h:99</div></div>
|
|
<div class="ttc" id="thread__multiply__add_8h_html"><div class="ttname"><a href="thread__multiply__add_8h.html">thread_multiply_add.h</a></div><div class="ttdoc">Template implementing matrix multiply-add operations on fragments. </div></div>
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4_html_a245bdf059794144d04d2823d2b39588e"><div class="ttname"><a href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a245bdf059794144d04d2823d2b39588e">cutlass::gemm::ThreadMultiplyAdd< ThreadGemmShape_, ThreadsPerWarp_, half, half, half >::InstructionShape</a></div><div class="ttdeci">Shape< 1, 1, 2, 1 > InstructionShape</div><div class="ttdoc">The shape of the instruction. </div><div class="ttdef"><b>Definition:</b> hgemm_multiply_add.h:44</div></div>
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4_html_aacb9a5a1d8f3b6e21bc449b0b97949b7"><div class="ttname"><a href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#aacb9a5a1d8f3b6e21bc449b0b97949b7">cutlass::gemm::ThreadMultiplyAdd< ThreadGemmShape_, ThreadsPerWarp_, half, half, half >::AccumulatorsPerWarp</a></div><div class="ttdeci">ShapeMul< ThreadGemmShape, ThreadsPerWarp >::Shape AccumulatorsPerWarp</div><div class="ttdoc">The number of accumulators per warp. </div><div class="ttdef"><b>Definition:</b> hgemm_multiply_add.h:52</div></div>
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4_html_a1889bdc9e88265a8afdaeeca217a3372"><div class="ttname"><a href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a1889bdc9e88265a8afdaeeca217a3372">cutlass::gemm::ThreadMultiplyAdd< ThreadGemmShape_, ThreadsPerWarp_, half, half, half >::FragmentA</a></div><div class="ttdeci">Fragment< ScalarA, AccumulatorsPerThread::kW > FragmentA</div><div class="ttdoc">The fragment for A. </div><div class="ttdef"><b>Definition:</b> hgemm_multiply_add.h:56</div></div>
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4_html_af8124acb485709dba1c5378faa24516c"><div class="ttname"><a href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#af8124acb485709dba1c5378faa24516c">cutlass::gemm::ThreadMultiplyAdd< ThreadGemmShape_, ThreadsPerWarp_, half, half, half >::Accumulators</a></div><div class="ttdeci">Fragment< half, AccumulatorsPerThread::kH *AccumulatorsPerThread::kW > Accumulators</div><div class="ttdoc">The accumulators. </div><div class="ttdef"><b>Definition:</b> hgemm_multiply_add.h:64</div></div>
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4_html_aaf8ea4aaf393488f03ba7fb7af264940"><div class="ttname"><a href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#aaf8ea4aaf393488f03ba7fb7af264940">cutlass::gemm::ThreadMultiplyAdd< ThreadGemmShape_, ThreadsPerWarp_, half, half, half >::multiply_add</a></div><div class="ttdeci">CUTLASS_DEVICE void multiply_add(FragmentA const &a, FragmentB const &b, Accumulators const &c, Accumulators &d)</div><div class="ttdoc">Multiply : d = a*b + c. </div><div class="ttdef"><b>Definition:</b> hgemm_multiply_add.h:74</div></div>
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4_html_a1a8f6feed85c7e88b36bc1a2637c716f"><div class="ttname"><a href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a1a8f6feed85c7e88b36bc1a2637c716f">cutlass::gemm::ThreadMultiplyAdd< ThreadGemmShape_, ThreadsPerWarp_, half, half, half >::ScalarA</a></div><div class="ttdeci">half ScalarA</div><div class="ttdoc">The type for A. </div><div class="ttdef"><b>Definition:</b> hgemm_multiply_add.h:54</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="structcutlass_1_1Shape_html"><div class="ttname"><a href="structcutlass_1_1Shape.html">cutlass::Shape</a></div><div class="ttdoc">A Shape implementing Layout Concept describing the dimensions of a cube. </div><div class="ttdef"><b>Definition:</b> shape.h:64</div></div>
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_html"><div class="ttname"><a href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd.html">cutlass::gemm::ThreadMultiplyAdd</a></div><div class="ttdoc">Template performing matrix multiply-add operation within a thread. </div><div class="ttdef"><b>Definition:</b> thread_multiply_add.h:44</div></div>
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4_html_a24dd9fdc54b001840e8b82664b3bfe3a"><div class="ttname"><a href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a24dd9fdc54b001840e8b82664b3bfe3a">cutlass::gemm::ThreadMultiplyAdd< ThreadGemmShape_, ThreadsPerWarp_, half, half, half >::AccumulatorsPerThread</a></div><div class="ttdeci">ThreadGemmShape AccumulatorsPerThread</div><div class="ttdoc">Aliased for compatibility. Will be removed for CUTLASS v2.0. </div><div class="ttdef"><b>Definition:</b> hgemm_multiply_add.h:48</div></div>
|
|
<div class="ttc" id="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4_html_a61ed3d8bb4ca6db39d16b632f58d75dc"><div class="ttname"><a href="structcutlass_1_1gemm_1_1ThreadMultiplyAdd_3_01ThreadGemmShape___00_01ThreadsPerWarp___00_01half_00_01half_00_01half_01_4.html#a61ed3d8bb4ca6db39d16b632f58d75dc">cutlass::gemm::ThreadMultiplyAdd< ThreadGemmShape_, ThreadsPerWarp_, half, half, half >::ThreadsPerWarp</a></div><div class="ttdeci">ThreadsPerWarp_ ThreadsPerWarp</div><div class="ttdoc">The number of threads per warp. </div><div class="ttdef"><b>Definition:</b> hgemm_multiply_add.h:50</div></div>
|
|
<div class="ttc" id="fragment_8h_html"><div class="ttname"><a href="fragment_8h.html">fragment.h</a></div><div class="ttdoc">Defines Fragment, a statically-sized array for storing parts of matrices within a thread&#39;s registers...</div></div>
|
|
</div><!-- fragment --></div><!-- contents -->
|
|
<!-- start footer part -->
|
|
<hr class="footer"/><address class="footer"><small>
|
|
Generated on Fri Oct 26 2018 14:53:33 for Cutlass by  <a href="http://www.doxygen.org/index.html">
|
|
<img class="footer" src="doxygen.png" alt="doxygen"/>
|
|
</a> 1.8.14
|
|
</small></address>
|
|
</body>
|
|
</html>
|