Google Git
Sign in
apache / mxnet-site / d5c5a235eef72f64ccaf224b834a20ef3287469b / . / versions / master / api / cpp / docs / api / group__dnnl__api__blas.html
blob: 01929a334e4c493bb1a4a0eca6d7502568e88268 [file] [log] [blame]
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "https://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 1.8.17"/>
<meta name="viewport" content="width=device-width, initial-scale=1"/>
<title>mxnet: Dnnl_api_blas</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>
<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">mxnet
</div>
</td>
</tr>
</tbody>
</table>
</div>
<!-- end header part -->
<!-- Generated by Doxygen 1.8.17 -->
<script type="text/javascript">
/* @license magnet:?xt=urn:btih:cf05388f2679ee054f2beb29a391d25f4e673ac3&amp;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&amp;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>
</div><!-- top -->
<!-- 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 class="header">
<div class="summary">
<a href="#func-members">Functions</a> </div>
<div class="headertitle">
<div class="title">Dnnl_api_blas<div class="ingroups"><a class="el" href="group__dnnl__api.html">Dnnl_api</a></div></div> </div>
</div><!--header-->
<div class="contents">
<div class="dynheader">
Collaboration diagram for Dnnl_api_blas:</div>
<div class="dyncontent">
<div class="center"><img src="group__dnnl__api__blas.png" border="0" usemap="#group____dnnl____api____blas" alt=""/></div>
<!-- MAP 0 -->
</div>
<table class="memberdecls">
<tr class="heading"><td colspan="2"><h2 class="groupheader"><a name="func-members"></a>
Functions</h2></td></tr>
<tr class="memitem:ga75ee119765bdac249200fda42c0617f8"><td class="memItemLeft" align="right" valign="top"><a class="el" href="group__dnnl__api__utils.html#gad24f9ded06e34d3ee71e7fc4b408d57a">dnnl_status_t</a> DNNL_API&#160;</td><td class="memItemRight" valign="bottom"><a class="el" href="group__dnnl__api__blas.html#ga75ee119765bdac249200fda42c0617f8">dnnl_sgemm</a> (char transa, char transb, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> M, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> N, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> K, float alpha, const float *A, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> lda, const float *B, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> ldb, float beta, float *C, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> ldc)</td></tr>
<tr class="separator:ga75ee119765bdac249200fda42c0617f8"><td class="memSeparator" colspan="2">&#160;</td></tr>
<tr class="memitem:gaef24848fd198d8a178d3ad95a78c1767"><td class="memItemLeft" align="right" valign="top"><a class="el" href="group__dnnl__api__utils.html#gad24f9ded06e34d3ee71e7fc4b408d57a">dnnl_status_t</a> DNNL_API&#160;</td><td class="memItemRight" valign="bottom"><a class="el" href="group__dnnl__api__blas.html#gaef24848fd198d8a178d3ad95a78c1767">dnnl_gemm_u8s8s32</a> (char transa, char transb, char offsetc, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> M, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> N, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> K, float alpha, const uint8_t *A, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> lda, uint8_t ao, const int8_t *B, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> ldb, int8_t bo, float beta, int32_t *C, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> ldc, const int32_t *co)</td></tr>
<tr class="separator:gaef24848fd198d8a178d3ad95a78c1767"><td class="memSeparator" colspan="2">&#160;</td></tr>
<tr class="memitem:ga2b763b7629846913507d88fba875cc26"><td class="memItemLeft" align="right" valign="top"><a class="el" href="group__dnnl__api__utils.html#gad24f9ded06e34d3ee71e7fc4b408d57a">dnnl_status_t</a> DNNL_API&#160;</td><td class="memItemRight" valign="bottom"><a class="el" href="group__dnnl__api__blas.html#ga2b763b7629846913507d88fba875cc26">dnnl_gemm_s8s8s32</a> (char transa, char transb, char offsetc, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> M, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> N, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> K, float alpha, const int8_t *A, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> lda, int8_t ao, const int8_t *B, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> ldb, int8_t bo, float beta, int32_t *C, <a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a> ldc, const int32_t *co)</td></tr>
<tr class="separator:ga2b763b7629846913507d88fba875cc26"><td class="memSeparator" colspan="2">&#160;</td></tr>
</table>
<a name="details" id="details"></a><h2 class="groupheader">Detailed Description</h2>
<h2 class="groupheader">Function Documentation</h2>
<a id="ga2b763b7629846913507d88fba875cc26"></a>
<h2 class="memtitle"><span class="permalink"><a href="#ga2b763b7629846913507d88fba875cc26">&#9670;&nbsp;</a></span>dnnl_gemm_s8s8s32()</h2>
<div class="memitem">
<div class="memproto">
<table class="memname">
<tr>
<td class="memname"><a class="el" href="group__dnnl__api__utils.html#gad24f9ded06e34d3ee71e7fc4b408d57a">dnnl_status_t</a> DNNL_API dnnl_gemm_s8s8s32 </td>
<td>(</td>
<td class="paramtype">char&#160;</td>
<td class="paramname"><em>transa</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">char&#160;</td>
<td class="paramname"><em>transb</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">char&#160;</td>
<td class="paramname"><em>offsetc</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>M</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>N</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>K</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">float&#160;</td>
<td class="paramname"><em>alpha</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">const int8_t *&#160;</td>
<td class="paramname"><em>A</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>lda</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">int8_t&#160;</td>
<td class="paramname"><em>ao</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">const int8_t *&#160;</td>
<td class="paramname"><em>B</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>ldb</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">int8_t&#160;</td>
<td class="paramname"><em>bo</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">float&#160;</td>
<td class="paramname"><em>beta</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">int32_t *&#160;</td>
<td class="paramname"><em>C</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>ldc</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">const int32_t *&#160;</td>
<td class="paramname"><em>co</em>&#160;</td>
</tr>
<tr>
<td></td>
<td>)</td>
<td></td><td></td>
</tr>
</table>
</div><div class="memdoc">
<p>Performs integer matrix-matrix multiply on 8-bit signed matrix A, 8-bit signed matrix B, and 32-bit signed resulting matrix C.</p>
<p>The operation is defined as:</p>
<p><code>C := alpha * (op(A) - A_offset) * (op(B) - B_offset) + beta * C + C_offset</code></p>
<p>where</p><ul>
<li><code>op( X ) = X</code> or <code>op( X ) = X**T</code>,</li>
<li><code>alpha</code> and <code>beta</code> are scalars, and</li>
<li><code>A</code>, <code>B</code>, and <code>C</code> are matrices:<ul>
<li><code>op( A )</code> is an <code>MxK</code> matrix,</li>
<li><code>op( B )</code> is an <code>KxN</code> matrix,</li>
<li><code>C</code> is an <code>MxN</code> matrix.</li>
</ul>
</li>
<li><code>A_offset</code> is an <code>MxK</code> matrix with every element equal the <code>ao</code> value,</li>
<li><code>B_offset</code> is an <code>KxN</code> matrix with every element equal the <code>bo</code> value,</li>
<li><code>C_offset</code> is an <code>MxN</code> matrix which is defined by the <code>co</code> array of size <code>len</code>:<ul>
<li>if <code>offsetc = F</code>: the <code>len</code> must be at least <code>1</code>,</li>
<li>if <code>offsetc = C</code>: the <code>len</code> must be at least <code>max(1, m)</code>,</li>
<li>if <code>offsetc = R</code>: the <code>len</code> must be at least <code>max(1, n)</code>,</li>
</ul>
</li>
</ul>
<p>The matrices are assumed to be stored in row-major order (the elements in each of the matrix rows are contiguous in memory).</p>
<dl class="section note"><dt>Note</dt><dd>This API does not support XERBLA. Instead, unlike the standard BLAS functions, this one returns a dnnl_status_t value to allow error handling.</dd></dl>
<dl class="section warning"><dt>Warning</dt><dd>On some architectures saturation may happen during intermediate computations, which would lead to unexpected results. For more details, refer to dev_guide_int8_computations.</dd></dl>
<dl class="params"><dt>Parameters</dt><dd>
<table class="params">
<tr><td class="paramname">transa</td><td>Transposition flag for matrix A: 'N' or 'n' means A is not transposed, and 'T' or 't' means that A is transposed. </td></tr>
<tr><td class="paramname">transb</td><td>Transposition flag for matrix B: 'N' or 'n' means B is not transposed, and 'T' or 't' means that B is transposed. </td></tr>
<tr><td class="paramname">offsetc</td><td>Flag specifying how offsets should be applied to matrix C:<ul>
<li>'F' means that the same offset will be applied to each element of the matrix C,</li>
<li>'C' means that individual offset will be applied to each element within each column,</li>
<li>'R' means that individual offset will be applied to each element within each row. </li>
</ul>
</td></tr>
<tr><td class="paramname">M</td><td>The M dimension. </td></tr>
<tr><td class="paramname">N</td><td>The N dimension. </td></tr>
<tr><td class="paramname">K</td><td>The K dimension. </td></tr>
<tr><td class="paramname">alpha</td><td>The alpha parameter that is used to scale the product of matrices A and B. </td></tr>
<tr><td class="paramname">A</td><td>A pointer to the A matrix data. </td></tr>
<tr><td class="paramname">lda</td><td>The leading dimension for the matrix A. </td></tr>
<tr><td class="paramname">ao</td><td>The offset value for the matrix A. </td></tr>
<tr><td class="paramname">B</td><td>A pointer to the B matrix data. </td></tr>
<tr><td class="paramname">ldb</td><td>The leading dimension for the matrix B. </td></tr>
<tr><td class="paramname">bo</td><td>The offset value for the matrix B. </td></tr>
<tr><td class="paramname">beta</td><td>The beta parameter that is used to scale the matrix C. </td></tr>
<tr><td class="paramname">C</td><td>A pointer to the C matrix data. </td></tr>
<tr><td class="paramname">ldc</td><td>The leading dimension for the matrix C. </td></tr>
<tr><td class="paramname">co</td><td>An array of offset values for the matrix C. The number of elements in the array depends on the value of <code>offsetc</code>. </td></tr>
</table>
</dd>
</dl>
<dl class="section return"><dt>Returns</dt><dd><a class="el" href="group__dnnl__api__utils.html#ggad24f9ded06e34d3ee71e7fc4b408d57aaa31395e9dccc103cf166cf7b38fc5b9c" title="The operation was successful.">dnnl_success</a>/#dnnl::status::success on success and a status describing the error otherwise. </dd></dl>
</div>
</div>
<a id="gaef24848fd198d8a178d3ad95a78c1767"></a>
<h2 class="memtitle"><span class="permalink"><a href="#gaef24848fd198d8a178d3ad95a78c1767">&#9670;&nbsp;</a></span>dnnl_gemm_u8s8s32()</h2>
<div class="memitem">
<div class="memproto">
<table class="memname">
<tr>
<td class="memname"><a class="el" href="group__dnnl__api__utils.html#gad24f9ded06e34d3ee71e7fc4b408d57a">dnnl_status_t</a> DNNL_API dnnl_gemm_u8s8s32 </td>
<td>(</td>
<td class="paramtype">char&#160;</td>
<td class="paramname"><em>transa</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">char&#160;</td>
<td class="paramname"><em>transb</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">char&#160;</td>
<td class="paramname"><em>offsetc</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>M</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>N</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>K</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">float&#160;</td>
<td class="paramname"><em>alpha</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">const uint8_t *&#160;</td>
<td class="paramname"><em>A</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>lda</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">uint8_t&#160;</td>
<td class="paramname"><em>ao</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">const int8_t *&#160;</td>
<td class="paramname"><em>B</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>ldb</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">int8_t&#160;</td>
<td class="paramname"><em>bo</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">float&#160;</td>
<td class="paramname"><em>beta</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">int32_t *&#160;</td>
<td class="paramname"><em>C</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>ldc</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">const int32_t *&#160;</td>
<td class="paramname"><em>co</em>&#160;</td>
</tr>
<tr>
<td></td>
<td>)</td>
<td></td><td></td>
</tr>
</table>
</div><div class="memdoc">
<p>Performs integer matrix-matrix multiply on 8-bit unsigned matrix A, 8-bit signed matrix B, and 32-bit signed resulting matrix C.</p>
<p>The operation is defined as:</p>
<p><code>C := alpha * (op(A) - A_offset) * (op(B) - B_offset) + beta * C + C_offset</code></p>
<p>where</p><ul>
<li><code>op( X ) = X</code> or <code>op( X ) = X**T</code>,</li>
<li><code>alpha</code> and <code>beta</code> are scalars, and</li>
<li><code>A</code>, <code>B</code>, and <code>C</code> are matrices:<ul>
<li><code>op( A )</code> is an <code>MxK</code> matrix,</li>
<li><code>op( B )</code> is an <code>KxN</code> matrix,</li>
<li><code>C</code> is an <code>MxN</code> matrix.</li>
</ul>
</li>
<li><code>A_offset</code> is an <code>MxK</code> matrix with every element equal the <code>ao</code> value,</li>
<li><code>B_offset</code> is an <code>KxN</code> matrix with every element equal the <code>bo</code> value,</li>
<li><code>C_offset</code> is an <code>MxN</code> matrix which is defined by the <code>co</code> array of size <code>len</code>:<ul>
<li>if <code>offsetc = F</code>: the <code>len</code> must be at least <code>1</code>,</li>
<li>if <code>offsetc = C</code>: the <code>len</code> must be at least <code>max(1, m)</code>,</li>
<li>if <code>offsetc = R</code>: the <code>len</code> must be at least <code>max(1, n)</code>,</li>
</ul>
</li>
</ul>
<p>The matrices are assumed to be stored in row-major order (the elements in each of the matrix rows are contiguous in memory).</p>
<dl class="section note"><dt>Note</dt><dd>This API does not support XERBLA. Instead, unlike the standard BLAS functions, this one returns a dnnl_status_t value to allow error handling.</dd></dl>
<dl class="section warning"><dt>Warning</dt><dd>On some architectures saturation may happen during intermediate computations, which would lead to unexpected results. For more details, refer to dev_guide_int8_computations.</dd></dl>
<dl class="params"><dt>Parameters</dt><dd>
<table class="params">
<tr><td class="paramname">transa</td><td>Transposition flag for matrix A: 'N' or 'n' means A is not transposed, and 'T' or 't' means that A is transposed. </td></tr>
<tr><td class="paramname">transb</td><td>Transposition flag for matrix B: 'N' or 'n' means B is not transposed, and 'T' or 't' means that B is transposed. </td></tr>
<tr><td class="paramname">offsetc</td><td>Flag specifying how offsets should be applied to matrix C:<ul>
<li>'F' means that the same offset will be applied to each element of the matrix C,</li>
<li>'C' means that individual offset will be applied to each element within each column,</li>
<li>'R' means that individual offset will be applied to each element within each row. </li>
</ul>
</td></tr>
<tr><td class="paramname">M</td><td>The M dimension. </td></tr>
<tr><td class="paramname">N</td><td>The N dimension. </td></tr>
<tr><td class="paramname">K</td><td>The K dimension. </td></tr>
<tr><td class="paramname">alpha</td><td>The alpha parameter that is used to scale the product of matrices A and B. </td></tr>
<tr><td class="paramname">A</td><td>A pointer to the A matrix data. </td></tr>
<tr><td class="paramname">lda</td><td>The leading dimension for the matrix A. </td></tr>
<tr><td class="paramname">ao</td><td>The offset value for the matrix A. </td></tr>
<tr><td class="paramname">B</td><td>A pointer to the B matrix data. </td></tr>
<tr><td class="paramname">ldb</td><td>The leading dimension for the matrix B. </td></tr>
<tr><td class="paramname">bo</td><td>The offset value for the matrix B. </td></tr>
<tr><td class="paramname">beta</td><td>The beta parameter that is used to scale the matrix C. </td></tr>
<tr><td class="paramname">C</td><td>A pointer to the C matrix data. </td></tr>
<tr><td class="paramname">ldc</td><td>The leading dimension for the matrix C. </td></tr>
<tr><td class="paramname">co</td><td>An array of offset values for the matrix C. The number of elements in the array depends on the value of <code>offsetc</code>. </td></tr>
</table>
</dd>
</dl>
<dl class="section return"><dt>Returns</dt><dd><a class="el" href="group__dnnl__api__utils.html#ggad24f9ded06e34d3ee71e7fc4b408d57aaa31395e9dccc103cf166cf7b38fc5b9c" title="The operation was successful.">dnnl_success</a>/#dnnl::status::success on success and a status describing the error otherwise. </dd></dl>
</div>
</div>
<a id="ga75ee119765bdac249200fda42c0617f8"></a>
<h2 class="memtitle"><span class="permalink"><a href="#ga75ee119765bdac249200fda42c0617f8">&#9670;&nbsp;</a></span>dnnl_sgemm()</h2>
<div class="memitem">
<div class="memproto">
<table class="memname">
<tr>
<td class="memname"><a class="el" href="group__dnnl__api__utils.html#gad24f9ded06e34d3ee71e7fc4b408d57a">dnnl_status_t</a> DNNL_API dnnl_sgemm </td>
<td>(</td>
<td class="paramtype">char&#160;</td>
<td class="paramname"><em>transa</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">char&#160;</td>
<td class="paramname"><em>transb</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>M</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>N</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>K</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">float&#160;</td>
<td class="paramname"><em>alpha</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">const float *&#160;</td>
<td class="paramname"><em>A</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>lda</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">const float *&#160;</td>
<td class="paramname"><em>B</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>ldb</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">float&#160;</td>
<td class="paramname"><em>beta</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype">float *&#160;</td>
<td class="paramname"><em>C</em>, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
<td class="paramtype"><a class="el" href="group__dnnl__api__memory.html#ga872631b12a112bf43fba985cba24dd20">dnnl_dim_t</a>&#160;</td>
<td class="paramname"><em>ldc</em>&#160;</td>
</tr>
<tr>
<td></td>
<td>)</td>
<td></td><td></td>
</tr>
</table>
</div><div class="memdoc">
<p>Performs single-precision matrix-matrix multiply.</p>
<p>The operation is defined as:</p>
<p><code>C := alpha * op( A ) * op( B ) + beta * C</code></p>
<p>where</p><ul>
<li><code>op( X ) = X</code> or <code>op( X ) = X**T</code>,</li>
<li><code>alpha</code> and <code>beta</code> are scalars, and</li>
<li><code>A</code>, <code>B</code>, and <code>C</code> are matrices:<ul>
<li><code>op( A )</code> is an <code>MxK</code> matrix,</li>
<li><code>op( B )</code> is an <code>KxN</code> matrix,</li>
<li><code>C</code> is an <code>MxN</code> matrix.</li>
</ul>
</li>
</ul>
<p>The matrices are assumed to be stored in row-major order (the elements in each of the matrix rows are contiguous in memory).</p>
<dl class="section note"><dt>Note</dt><dd>This API does not support XERBLA. Instead, unlike the standard BLAS functions, this one returns a dnnl_status_t value to allow error handling.</dd></dl>
<dl class="params"><dt>Parameters</dt><dd>
<table class="params">
<tr><td class="paramname">transa</td><td>Transposition flag for matrix A: 'N' or 'n' means A is not transposed, and 'T' or 't' means that A is transposed. </td></tr>
<tr><td class="paramname">transb</td><td>Transposition flag for matrix B: 'N' or 'n' means B is not transposed, and 'T' or 't' means that B is transposed. </td></tr>
<tr><td class="paramname">M</td><td>The M dimension. </td></tr>
<tr><td class="paramname">N</td><td>The N dimension. </td></tr>
<tr><td class="paramname">K</td><td>The K dimension. </td></tr>
<tr><td class="paramname">alpha</td><td>The alpha parameter that is used to scale the product of matrices A and B. </td></tr>
<tr><td class="paramname">A</td><td>A pointer to the A matrix data. </td></tr>
<tr><td class="paramname">lda</td><td>The leading dimension for the matrix A. </td></tr>
<tr><td class="paramname">B</td><td>A pointer to the B matrix data. </td></tr>
<tr><td class="paramname">ldb</td><td>The leading dimension for the matrix B. </td></tr>
<tr><td class="paramname">beta</td><td>The beta parameter that is used to scale the matrix C. </td></tr>
<tr><td class="paramname">C</td><td>A pointer to the C matrix data. </td></tr>
<tr><td class="paramname">ldc</td><td>The leading dimension for the matrix C. </td></tr>
</table>
</dd>
</dl>
<dl class="section return"><dt>Returns</dt><dd><a class="el" href="group__dnnl__api__utils.html#ggad24f9ded06e34d3ee71e7fc4b408d57aaa31395e9dccc103cf166cf7b38fc5b9c" title="The operation was successful.">dnnl_success</a>/#dnnl::status::success on success and a status describing the error otherwise. </dd></dl>
</div>
</div>
</div><!-- contents -->
<!-- start footer part -->
<hr class="footer"/><address class="footer"><small>
Generated on Thu Jan 5 2023 03:47:40 for mxnet by &#160;<a href="http://www.doxygen.org/index.html">
<img class="footer" src="doxygen.png" alt="doxygen"/>
</a> 1.8.17
</small></address>
</body>
</html>