dnrops
/
cutlass
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
cutlass/docs/platform_8h_source.html

183 lines
149 KiB
HTML
Raw Permalink Normal View History

CUTLASS v1.0 release
2018-05-17 02:44:56 +08:00
<!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"/>
CUTLASS 2.0 (#62) CUTLASS 2.0 Substantially refactored for - Better performance, particularly for native Turing Tensor Cores - Robust and durable templates spanning the design space - Encapsulated functionality embodying modern C++11 programming techniques - Optimized containers and data types for efficient, generic, portable device code Updates to: - Quick start guide - Documentation - Utilities - CUTLASS Profiler Native Turing Tensor Cores - Efficient GEMM kernels targeting Turing Tensor Cores - Mixed-precision floating point, 8-bit integer, 4-bit integer, and binarized operands Coverage of existing CUTLASS functionality: - GEMM kernels targeting CUDA and Tensor Cores in NVIDIA GPUs - Volta Tensor Cores through native mma.sync and through WMMA API - Optimizations such as parallel reductions, threadblock rasterization, and intra-threadblock reductions - Batched GEMM operations - Complex-valued GEMMs Note: this commit and all that follow require a host compiler supporting C++11 or greater.
2019-11-20 08:55:34 +08:00
<meta name="generator" content="Doxygen 1.8.11"/>
<title>CUTLASS: platform.h Source File</title>
CUTLASS v1.0 release
2018-05-17 02:44:56 +08:00
<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>
CUTLASS 2.0 (#62) CUTLASS 2.0 Substantially refactored for - Better performance, particularly for native Turing Tensor Cores - Robust and durable templates spanning the design space - Encapsulated functionality embodying modern C++11 programming techniques - Optimized containers and data types for efficient, generic, portable device code Updates to: - Quick start guide - Documentation - Utilities - CUTLASS Profiler Native Turing Tensor Cores - Efficient GEMM kernels targeting Turing Tensor Cores - Mixed-precision floating point, 8-bit integer, 4-bit integer, and binarized operands Coverage of existing CUTLASS functionality: - GEMM kernels targeting CUDA and Tensor Cores in NVIDIA GPUs - Volta Tensor Cores through native mma.sync and through WMMA API - Optimizations such as parallel reductions, threadblock rasterization, and intra-threadblock reductions - Batched GEMM operations - Complex-valued GEMMs Note: this commit and all that follow require a host compiler supporting C++11 or greater.
2019-11-20 08:55:34 +08:00
<script type="text/javascript">
$(document).ready(function() { init_search(); });
</script>
CUTLASS v1.0 release
2018-05-17 02:44:56 +08:00
<script type="text/x-mathjax-config">
MathJax.Hub.Config({
extensions: ["tex2jax.js"],
jax: ["input/TeX","output/HTML-CSS"],
});
CUTLASS 2.0 (#62) CUTLASS 2.0 Substantially refactored for - Better performance, particularly for native Turing Tensor Cores - Robust and durable templates spanning the design space - Encapsulated functionality embodying modern C++11 programming techniques - Optimized containers and data types for efficient, generic, portable device code Updates to: - Quick start guide - Documentation - Utilities - CUTLASS Profiler Native Turing Tensor Cores - Efficient GEMM kernels targeting Turing Tensor Cores - Mixed-precision floating point, 8-bit integer, 4-bit integer, and binarized operands Coverage of existing CUTLASS functionality: - GEMM kernels targeting CUDA and Tensor Cores in NVIDIA GPUs - Volta Tensor Cores through native mma.sync and through WMMA API - Optimizations such as parallel reductions, threadblock rasterization, and intra-threadblock reductions - Batched GEMM operations - Complex-valued GEMMs Note: this commit and all that follow require a host compiler supporting C++11 or greater.
2019-11-20 08:55:34 +08:00
</script><script type="text/javascript" src="http://cdn.mathjax.org/mathjax/latest/MathJax.js"></script>
CUTLASS v1.0 release
2018-05-17 02:44:56 +08:00
<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;">
CUTLASS 2.0 (#62) CUTLASS 2.0 Substantially refactored for - Better performance, particularly for native Turing Tensor Cores - Robust and durable templates spanning the design space - Encapsulated functionality embodying modern C++11 programming techniques - Optimized containers and data types for efficient, generic, portable device code Updates to: - Quick start guide - Documentation - Utilities - CUTLASS Profiler Native Turing Tensor Cores - Efficient GEMM kernels targeting Turing Tensor Cores - Mixed-precision floating point, 8-bit integer, 4-bit integer, and binarized operands Coverage of existing CUTLASS functionality: - GEMM kernels targeting CUDA and Tensor Cores in NVIDIA GPUs - Volta Tensor Cores through native mma.sync and through WMMA API - Optimizations such as parallel reductions, threadblock rasterization, and intra-threadblock reductions - Batched GEMM operations - Complex-valued GEMMs Note: this commit and all that follow require a host compiler supporting C++11 or greater.
2019-11-20 08:55:34 +08:00
<td id="projectlogo"><img alt="Logo" src="cutlass-logo-small.png"/></td>
CUTLASS v1.0 release
2018-05-17 02:44:56 +08:00
<td id="projectalign" style="padding-left: 0.5em;">
CUTLASS 2.0 (#62) CUTLASS 2.0 Substantially refactored for - Better performance, particularly for native Turing Tensor Cores - Robust and durable templates spanning the design space - Encapsulated functionality embodying modern C++11 programming techniques - Optimized containers and data types for efficient, generic, portable device code Updates to: - Quick start guide - Documentation - Utilities - CUTLASS Profiler Native Turing Tensor Cores - Efficient GEMM kernels targeting Turing Tensor Cores - Mixed-precision floating point, 8-bit integer, 4-bit integer, and binarized operands Coverage of existing CUTLASS functionality: - GEMM kernels targeting CUDA and Tensor Cores in NVIDIA GPUs - Volta Tensor Cores through native mma.sync and through WMMA API - Optimizations such as parallel reductions, threadblock rasterization, and intra-threadblock reductions - Batched GEMM operations - Complex-valued GEMMs Note: this commit and all that follow require a host compiler supporting C++11 or greater.
2019-11-20 08:55:34 +08:00
<div id="projectname">CUTLASS
CUTLASS v1.0 release
2018-05-17 02:44:56 +08:00
</div>
<div id="projectbrief">CUDA Templates for Linear Algebra Subroutines and Solvers</div>
</td>
</tr>
</tbody>
</table>
</div>
<!-- end header part -->
CUTLASS 2.0 (#62) CUTLASS 2.0 Substantially refactored for - Better performance, particularly for native Turing Tensor Cores - Robust and durable templates spanning the design space - Encapsulated functionality embodying modern C++11 programming techniques - Optimized containers and data types for efficient, generic, portable device code Updates to: - Quick start guide - Documentation - Utilities - CUTLASS Profiler Native Turing Tensor Cores - Efficient GEMM kernels targeting Turing Tensor Cores - Mixed-precision floating point, 8-bit integer, 4-bit integer, and binarized operands Coverage of existing CUTLASS functionality: - GEMM kernels targeting CUDA and Tensor Cores in NVIDIA GPUs - Volta Tensor Cores through native mma.sync and through WMMA API - Optimizations such as parallel reductions, threadblock rasterization, and intra-threadblock reductions - Batched GEMM operations - Complex-valued GEMMs Note: this commit and all that follow require a host compiler supporting C++11 or greater.
2019-11-20 08:55:34 +08:00
<!-- Generated by Doxygen 1.8.11 -->
CUTLASS v1.0 release
2018-05-17 02:44:56 +08:00
<script type="text/javascript">
var searchBox = new SearchBox("searchBox", "search",false,'Search');
</script>
CUTLASS 2.0 (#62) CUTLASS 2.0 Substantially refactored for - Better performance, particularly for native Turing Tensor Cores - Robust and durable templates spanning the design space - Encapsulated functionality embodying modern C++11 programming techniques - Optimized containers and data types for efficient, generic, portable device code Updates to: - Quick start guide - Documentation - Utilities - CUTLASS Profiler Native Turing Tensor Cores - Efficient GEMM kernels targeting Turing Tensor Cores - Mixed-precision floating point, 8-bit integer, 4-bit integer, and binarized operands Coverage of existing CUTLASS functionality: - GEMM kernels targeting CUDA and Tensor Cores in NVIDIA GPUs - Volta Tensor Cores through native mma.sync and through WMMA API - Optimizations such as parallel reductions, threadblock rasterization, and intra-threadblock reductions - Batched GEMM operations - Complex-valued GEMMs Note: this commit and all that follow require a host compiler supporting C++11 or greater.
2019-11-20 08:55:34 +08:00
<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>
CUTLASS v1.0 release
2018-05-17 02:44:56 +08:00
<!-- 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>
CUTLASS 2.0 (#62) CUTLASS 2.0 Substantially refactored for - Better performance, particularly for native Turing Tensor Cores - Robust and durable templates spanning the design space - Encapsulated functionality embodying modern C++11 programming techniques - Optimized containers and data types for efficient, generic, portable device code Updates to: - Quick start guide - Documentation - Utilities - CUTLASS Profiler Native Turing Tensor Cores - Efficient GEMM kernels targeting Turing Tensor Cores - Mixed-precision floating point, 8-bit integer, 4-bit integer, and binarized operands Coverage of existing CUTLASS functionality: - GEMM kernels targeting CUDA and Tensor Cores in NVIDIA GPUs - Volta Tensor Cores through native mma.sync and through WMMA API - Optimizations such as parallel reductions, threadblock rasterization, and intra-threadblock reductions - Batched GEMM operations - Complex-valued GEMMs Note: this commit and all that follow require a host compiler supporting C++11 or greater.
2019-11-20 08:55:34 +08:00
<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_df998829b150afe92f54393d2430470d.html">platform</a></li> </ul>
CUTLASS v1.0 release
2018-05-17 02:44:56 +08:00
</div>
</div><!-- top -->
<div class="header">
<div class="headertitle">
<div class="title">platform.h</div> </div>
</div><!--header-->
<div class="contents">
CUTLASS 2.0 (#62) CUTLASS 2.0 Substantially refactored for - Better performance, particularly for native Turing Tensor Cores - Robust and durable templates spanning the design space - Encapsulated functionality embodying modern C++11 programming techniques - Optimized containers and data types for efficient, generic, portable device code Updates to: - Quick start guide - Documentation - Utilities - CUTLASS Profiler Native Turing Tensor Cores - Efficient GEMM kernels targeting Turing Tensor Cores - Mixed-precision floating point, 8-bit integer, 4-bit integer, and binarized operands Coverage of existing CUTLASS functionality: - GEMM kernels targeting CUDA and Tensor Cores in NVIDIA GPUs - Volta Tensor Cores through native mma.sync and through WMMA API - Optimizations such as parallel reductions, threadblock rasterization, and intra-threadblock reductions - Batched GEMM operations - Complex-valued GEMMs Note: this commit and all that follow require a host compiler supporting C++11 or greater.
2019-11-20 08:55:34 +08:00
<a href="platform_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
<div class="ttc" id="namespacecutlass_1_1platform_html_af6a9a165e53d7e85ae121d5789aa03e0"><div class="ttname"><a href="namespacecutlass_1_1platform.html#af6a9a165e53d7e85ae121d5789aa03e0">cutlass::platform::max</a></div><div class="ttdeci">CUTLASS_HOST_DEVICE constexpr const T &amp; max(const T &amp;a, const T &amp;b)</div><div class="ttdoc">std::max </div><div class="ttdef"><b>Definition:</b> platform.h:189</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>
Updated doxygen
2018-09-20 05:02:08 +08:00
<div class="ttc" id="platform_8h_html_a72f0657181cca64b44eb186b707eb380"><div class="ttname"><a href="platform_8h.html#a72f0657181cca64b44eb186b707eb380">constexpr</a></div><div class="ttdeci">#define constexpr</div><div class="ttdef"><b>Definition:</b> platform.h:137</div></div>
CUTLASS 2.0 (#62) CUTLASS 2.0 Substantially refactored for - Better performance, particularly for native Turing Tensor Cores - Robust and durable templates spanning the design space - Encapsulated functionality embodying modern C++11 programming techniques - Optimized containers and data types for efficient, generic, portable device code Updates to: - Quick start guide - Documentation - Utilities - CUTLASS Profiler Native Turing Tensor Cores - Efficient GEMM kernels targeting Turing Tensor Cores - Mixed-precision floating point, 8-bit integer, 4-bit integer, and binarized operands Coverage of existing CUTLASS functionality: - GEMM kernels targeting CUDA and Tensor Cores in NVIDIA GPUs - Volta Tensor Cores through native mma.sync and through WMMA API - Optimizations such as parallel reductions, threadblock rasterization, and intra-threadblock reductions - Batched GEMM operations - Complex-valued GEMMs Note: this commit and all that follow require a host compiler supporting C++11 or greater.
2019-11-20 08:55:34 +08:00
<div class="ttc" id="structcutlass_1_1platform_1_1nullptr__t_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1nullptr__t.html">cutlass::platform::nullptr_t</a></div><div class="ttdoc">std::nullptr_t </div><div class="ttdef"><b>Definition:</b> platform.h:299</div></div>
<div class="ttc" id="namespacecutlass_1_1platform_html_a3e83320a39137d92042eb0bf93be9678"><div class="ttname"><a href="namespacecutlass_1_1platform.html#a3e83320a39137d92042eb0bf93be9678">cutlass::platform::swap</a></div><div class="ttdeci">void swap(unique_ptr&lt; T, Deleter &gt; &amp;lhs, unique_ptr&lt; T, Deleter &gt; &amp;rhs) noexcept</div><div class="ttdoc">Specializes the swap algorithm. </div><div class="ttdef"><b>Definition:</b> platform.h:777</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1is__pointer__helper_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1is__pointer__helper.html">cutlass::platform::is_pointer_helper</a></div><div class="ttdoc">Helper for std::is_pointer (false specialization) </div><div class="ttdef"><b>Definition:</b> platform.h:450</div></div>
<div class="ttc" id="classcutlass_1_1platform_1_1unique__ptr_html_a85cab9945c36dc56bd7d6adf30c0d252"><div class="ttname"><a href="classcutlass_1_1platform_1_1unique__ptr.html#a85cab9945c36dc56bd7d6adf30c0d252">cutlass::platform::unique_ptr::deleter_type</a></div><div class="ttdeci">Deleter deleter_type</div><div class="ttdef"><b>Definition:</b> platform.h:716</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1remove__const_html_ac3662947fa50251daf58240a9c798085"><div class="ttname"><a href="structcutlass_1_1platform_1_1remove__const.html#ac3662947fa50251daf58240a9c798085">cutlass::platform::remove_const::type</a></div><div class="ttdeci">T type</div><div class="ttdef"><b>Definition:</b> platform.h:351</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1alignment__of_1_1pad_html_abc729cc51d5c90b1d7b0df3092d47cd4"><div class="ttname"><a href="structcutlass_1_1platform_1_1alignment__of_1_1pad.html#abc729cc51d5c90b1d7b0df3092d47cd4">cutlass::platform::alignment_of::pad::val</a></div><div class="ttdeci">value_t val</div><div class="ttdef"><b>Definition:</b> platform.h:560</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1conditional_html_ab6484d0dd6449b5195c4e868026fed11"><div class="ttname"><a href="structcutlass_1_1platform_1_1conditional.html#ab6484d0dd6449b5195c4e868026fed11">cutlass::platform::conditional::type</a></div><div class="ttdeci">T type</div><div class="ttdef"><b>Definition:</b> platform.h:326</div></div>
<div class="ttc" id="classcutlass_1_1platform_1_1unique__ptr_html_ab6ce60d03d11b269c1e151dfa7c696f9"><div class="ttname"><a href="classcutlass_1_1platform_1_1unique__ptr.html#ab6ce60d03d11b269c1e151dfa7c696f9">cutlass::platform::unique_ptr::pointer</a></div><div class="ttdeci">T * pointer</div><div class="ttdef"><b>Definition:</b> platform.h:714</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_1platform_1_1is__pointer_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1is__pointer.html">cutlass::platform::is_pointer</a></div><div class="ttdoc">std::is_pointer </div><div class="ttdef"><b>Definition:</b> platform.h:458</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1integral__constant_html_ab2ed0b3506818139f1f96639742e79fd"><div class="ttname"><a href="structcutlass_1_1platform_1_1integral__constant.html#ab2ed0b3506818139f1f96639742e79fd">cutlass::platform::integral_constant::value_type</a></div><div class="ttdeci">value_t value_type</div><div class="ttdef"><b>Definition:</b> platform.h:263</div></div>
<div class="ttc" id="namespacecutlass_1_1platform_html_a90ce74c7faa4e27c888ce56e957b73d5"><div class="ttname"><a href="namespacecutlass_1_1platform.html#a90ce74c7faa4e27c888ce56e957b73d5">cutlass::platform::make_pair</a></div><div class="ttdeci">CUTLASS_HOST_DEVICE std::pair&lt; T1, T2 &gt; make_pair(T1 t, T2 u)</div><div class="ttdef"><b>Definition:</b> platform.h:232</div></div>
<div class="ttc" id="classcutlass_1_1platform_1_1unique__ptr_html_aa8a370bc7e4c2d99eb85e7fea27b3179"><div class="ttname"><a href="classcutlass_1_1platform_1_1unique__ptr.html#aa8a370bc7e4c2d99eb85e7fea27b3179">cutlass::platform::unique_ptr::unique_ptr</a></div><div class="ttdeci">unique_ptr()</div><div class="ttdef"><b>Definition:</b> platform.h:726</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1default__delete_3_01T[]_4_html_a5db6aa72c16a39437c9de615fa163f06"><div class="ttname"><a href="structcutlass_1_1platform_1_1default__delete_3_01T[]_4.html#a5db6aa72c16a39437c9de615fa163f06">cutlass::platform::default_delete&lt; T[]&gt;::operator()</a></div><div class="ttdeci">void operator()(T *ptr) const </div><div class="ttdef"><b>Definition:</b> platform.h:707</div></div>
<div class="ttc" id="namespacecutlass_1_1platform_html_ab9b8306ae9dc21fa646c49b68fa8e197"><div class="ttname"><a href="namespacecutlass_1_1platform.html#ab9b8306ae9dc21fa646c49b68fa8e197">cutlass::platform::operator==</a></div><div class="ttdeci">CUTLASS_HOST_DEVICE constexpr bool operator==(const pair&lt; T1, T2 &gt; &amp;lhs, const pair&lt; T1, T2 &gt; &amp;rhs)</div><div class="ttdef"><b>Definition:</b> platform.h:201</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1is__void_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1is__void.html">cutlass::platform::is_void</a></div><div class="ttdoc">std::is_void </div><div class="ttdef"><b>Definition:</b> platform.h:462</div></div>
<div class="ttc" id="namespacecutlass_1_1platform_html_ab0f21e67c0a4b5c6952042b502c6816f"><div class="ttname"><a href="namespacecutlass_1_1platform.html#ab0f21e67c0a4b5c6952042b502c6816f">cutlass::platform::operator&gt;=</a></div><div class="ttdeci">CUTLASS_HOST_DEVICE constexpr bool operator&gt;=(const pair&lt; T1, T2 &gt; &amp;lhs, const pair&lt; T1, T2 &gt; &amp;rhs)</div><div class="ttdef"><b>Definition:</b> platform.h:227</div></div>
<div class="ttc" id="classcutlass_1_1platform_1_1unique__ptr_html_afa52edcaef23461ce1f9c1dac349c24b"><div class="ttname"><a href="classcutlass_1_1platform_1_1unique__ptr.html#afa52edcaef23461ce1f9c1dac349c24b">cutlass::platform::unique_ptr::operator-&gt;</a></div><div class="ttdeci">pointer operator-&gt;() const noexcept</div><div class="ttdoc">Returns a pointer to the managed object. </div><div class="ttdef"><b>Definition:</b> platform.h:769</div></div>
<div class="ttc" id="namespacecutlass_1_1platform_html_a9e8e698d40b8df881991fde9ba2a1b12"><div class="ttname"><a href="namespacecutlass_1_1platform.html#a9e8e698d40b8df881991fde9ba2a1b12">cutlass::platform::operator&gt;</a></div><div class="ttdeci">CUTLASS_HOST_DEVICE constexpr bool operator&gt;(const pair&lt; T1, T2 &gt; &amp;lhs, const pair&lt; T1, T2 &gt; &amp;rhs)</div><div class="ttdef"><b>Definition:</b> platform.h:222</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1default__delete_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1default__delete.html">cutlass::platform::default_delete</a></div><div class="ttdoc">Default deleter. </div><div class="ttdef"><b>Definition:</b> platform.h:700</div></div>
<div class="ttc" id="namespacecutlass_1_1platform_html_a248f49adf09654d2cd04bd2760ab2566"><div class="ttname"><a href="namespacecutlass_1_1platform.html#a248f49adf09654d2cd04bd2760ab2566">cutlass::platform::operator!=</a></div><div class="ttdeci">CUTLASS_HOST_DEVICE constexpr bool operator!=(const pair&lt; T1, T2 &gt; &amp;lhs, const pair&lt; T1, T2 &gt; &amp;rhs)</div><div class="ttdef"><b>Definition:</b> platform.h:206</div></div>
<div class="ttc" id="classcutlass_1_1platform_1_1unique__ptr_html"><div class="ttname"><a href="classcutlass_1_1platform_1_1unique__ptr.html">cutlass::platform::unique_ptr</a></div><div class="ttdoc">std::unique_ptr </div><div class="ttdef"><b>Definition:</b> platform.h:712</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1alignment__of_1_1pad_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1alignment__of_1_1pad.html">cutlass::platform::alignment_of::pad</a></div><div class="ttdef"><b>Definition:</b> platform.h:559</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1is__floating__point_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1is__floating__point.html">cutlass::platform::is_floating_point</a></div><div class="ttdoc">std::is_floating_point </div><div class="ttdef"><b>Definition:</b> platform.h:498</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1conditional_3_01false_00_01T_00_01F_01_4_html_a8d55f500f667de560650554e9c220644"><div class="ttname"><a href="structcutlass_1_1platform_1_1conditional_3_01false_00_01T_00_01F_01_4.html#a8d55f500f667de560650554e9c220644">cutlass::platform::conditional&lt; false, T, F &gt;::type</a></div><div class="ttdeci">F type</div><div class="ttdef"><b>Definition:</b> platform.h:332</div></div>
<div class="ttc" id="namespacecutlass_1_1platform_html_ad8c95b2109070847b13d355120344380"><div class="ttname"><a href="namespacecutlass_1_1platform.html#ad8c95b2109070847b13d355120344380">cutlass::platform::false_type</a></div><div class="ttdeci">integral_constant&lt; bool, false &gt; false_type</div><div class="ttdoc">The type used as a compile-time boolean with false value. </div><div class="ttdef"><b>Definition:</b> platform.h:282</div></div>
<div class="ttc" id="classcutlass_1_1platform_1_1unique__ptr_html_aa427ab4ea4f2336ac6db28d53a4c11ac"><div class="ttname"><a href="classcutlass_1_1platform_1_1unique__ptr.html#aa427ab4ea4f2336ac6db28d53a4c11ac">cutlass::platform::unique_ptr::get_deleter</a></div><div class="ttdeci">Deleter const &amp; get_deleter() const noexcept</div><div class="ttdoc">Returns the deleter object. </div><div class="ttdef"><b>Definition:</b> platform.h:760</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1remove__cv_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1remove__cv.html">cutlass::platform::remove_cv</a></div><div class="ttdoc">std::remove_cv </div><div class="ttdef"><b>Definition:</b> platform.h:374</div></div>
<div class="ttc" id="classcutlass_1_1platform_1_1unique__ptr_html_a8902399dac4ab64f08f909f2ad9d4bcf"><div class="ttname"><a href="classcutlass_1_1platform_1_1unique__ptr.html#a8902399dac4ab64f08f909f2ad9d4bcf">cutlass::platform::unique_ptr::~unique_ptr</a></div><div class="ttdeci">~unique_ptr()</div><div class="ttdef"><b>Definition:</b> platform.h:729</div></div>
<div class="ttc" id="namespacecutlass_1_1platform_html_ac9068e2d027ffdf5cd564deecc2cb9e8"><div class="ttname"><a href="namespacecutlass_1_1platform.html#ac9068e2d027ffdf5cd564deecc2cb9e8">cutlass::platform::__align__</a></div><div class="ttdeci">struct __align__(1) aligned_chunk&lt; 1 &gt;</div><div class="ttdef"><b>Definition:</b> platform.h:633</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1remove__const_3_01const_01T_01_4_html_af68706cfaa6af14edc26ad5b974b47e3"><div class="ttname"><a href="structcutlass_1_1platform_1_1remove__const_3_01const_01T_01_4.html#af68706cfaa6af14edc26ad5b974b47e3">cutlass::platform::remove_const&lt; const T &gt;::type</a></div><div class="ttdeci">T type</div><div class="ttdef"><b>Definition:</b> platform.h:357</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1remove__volatile_html_a4f5b043d46206248d1bbbcf650707dd1"><div class="ttname"><a href="structcutlass_1_1platform_1_1remove__volatile.html#a4f5b043d46206248d1bbbcf650707dd1">cutlass::platform::remove_volatile::type</a></div><div class="ttdeci">T type</div><div class="ttdef"><b>Definition:</b> platform.h:363</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1is__integral_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1is__integral.html">cutlass::platform::is_integral</a></div><div class="ttdoc">std::is_integral </div><div class="ttdef"><b>Definition:</b> platform.h:466</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1is__trivially__copyable_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1is__trivially__copyable.html">cutlass::platform::is_trivially_copyable</a></div><div class="ttdef"><b>Definition:</b> platform.h:541</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1is__base__of__helper_1_1dummy_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1is__base__of__helper_1_1dummy.html">cutlass::platform::is_base_of_helper::dummy</a></div><div class="ttdef"><b>Definition:</b> platform.h:407</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1integral__constant_html_af58810ccead8f16ed88cd6a4afdc6e52"><div class="ttname"><a href="structcutlass_1_1platform_1_1integral__constant.html#af58810ccead8f16ed88cd6a4afdc6e52">cutlass::platform::integral_constant::type</a></div><div class="ttdeci">integral_constant&lt; value_t, V &gt; type</div><div class="ttdef"><b>Definition:</b> platform.h:264</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1is__arithmetic_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1is__arithmetic.html">cutlass::platform::is_arithmetic</a></div><div class="ttdoc">std::is_arithmetic </div><div class="ttdef"><b>Definition:</b> platform.h:505</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1alignment__of_1_1pad_html_a86f075f91b80918e968951713430f0b4"><div class="ttname"><a href="structcutlass_1_1platform_1_1alignment__of_1_1pad.html#a86f075f91b80918e968951713430f0b4">cutlass::platform::alignment_of::pad::byte</a></div><div class="ttdeci">char byte</div><div class="ttdef"><b>Definition:</b> platform.h:561</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1integral__constant_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1integral__constant.html">cutlass::platform::integral_constant</a></div><div class="ttdoc">std::integral_constant </div><div class="ttdef"><b>Definition:</b> platform.h:256</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1is__base__of_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1is__base__of.html">cutlass::platform::is_base_of</a></div><div class="ttdoc">std::is_base_of </div><div class="ttdef"><b>Definition:</b> platform.h:422</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1enable__if_html_aff9c0f270020cf097addf77e53a5af99"><div class="ttname"><a href="structcutlass_1_1platform_1_1enable__if.html#aff9c0f270020cf097addf77e53a5af99">cutlass::platform::enable_if::type</a></div><div class="ttdeci">T type</div><div class="ttdef"><b>Definition:</b> platform.h:316</div></div>
Updated doxygen
2018-09-20 05:02:08 +08:00
<div class="ttc" id="platform_8h_html_ab979d9d4b4923f7c54d6caa6e1a61936"><div class="ttname"><a href="platform_8h.html#ab979d9d4b4923f7c54d6caa6e1a61936">nullptr</a></div><div class="ttdeci">#define nullptr</div><div class="ttdoc">nullptr </div><div class="ttdef"><b>Definition:</b> platform.h:144</div></div>
CUTLASS 2.0 (#62) CUTLASS 2.0 Substantially refactored for - Better performance, particularly for native Turing Tensor Cores - Robust and durable templates spanning the design space - Encapsulated functionality embodying modern C++11 programming techniques - Optimized containers and data types for efficient, generic, portable device code Updates to: - Quick start guide - Documentation - Utilities - CUTLASS Profiler Native Turing Tensor Cores - Efficient GEMM kernels targeting Turing Tensor Cores - Mixed-precision floating point, 8-bit integer, 4-bit integer, and binarized operands Coverage of existing CUTLASS functionality: - GEMM kernels targeting CUDA and Tensor Cores in NVIDIA GPUs - Volta Tensor Cores through native mma.sync and through WMMA API - Optimizations such as parallel reductions, threadblock rasterization, and intra-threadblock reductions - Batched GEMM operations - Complex-valued GEMMs Note: this commit and all that follow require a host compiler supporting C++11 or greater.
2019-11-20 08:55:34 +08:00
<div class="ttc" id="structcutlass_1_1platform_1_1is__volatile_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1is__volatile.html">cutlass::platform::is_volatile</a></div><div class="ttdoc">std::is_volatile </div><div class="ttdef"><b>Definition:</b> platform.h:444</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1is__fundamental_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1is__fundamental.html">cutlass::platform::is_fundamental</a></div><div class="ttdoc">std::is_fundamental </div><div class="ttdef"><b>Definition:</b> platform.h:510</div></div>
<div class="ttc" id="classcutlass_1_1platform_1_1unique__ptr_html_a0bd8240db28f3964714da5113cd9e354"><div class="ttname"><a href="classcutlass_1_1platform_1_1unique__ptr.html#a0bd8240db28f3964714da5113cd9e354">cutlass::platform::unique_ptr::operator*</a></div><div class="ttdeci">T &amp; operator*() const </div><div class="ttdoc">Dereferences the unique_ptr. </div><div class="ttdef"><b>Definition:</b> platform.h:766</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1enable__if_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1enable__if.html">cutlass::platform::enable_if</a></div><div class="ttdoc">std::enable_if (true specialization) </div><div class="ttdef"><b>Definition:</b> platform.h:315</div></div>
<div class="ttc" id="namespacecutlass_1_1platform_html_a0eddc4a3921e137f31fd8014be96e807"><div class="ttname"><a href="namespacecutlass_1_1platform.html#a0eddc4a3921e137f31fd8014be96e807">cutlass::platform::true_type</a></div><div class="ttdeci">integral_constant&lt; bool, true &gt; true_type</div><div class="ttdoc">The type used as a compile-time boolean with true value. </div><div class="ttdef"><b>Definition:</b> platform.h:279</div></div>
<div class="ttc" id="cutlass_8h_html_a28c2443a142676d3d71effdae1a986b1"><div class="ttname"><a href="cutlass_8h.html#a28c2443a142676d3d71effdae1a986b1">CUTLASS_HOST_DEVICE</a></div><div class="ttdeci">#define CUTLASS_HOST_DEVICE</div><div class="ttdef"><b>Definition:</b> cutlass.h:89</div></div>
<div class="ttc" id="classcutlass_1_1platform_1_1unique__ptr_html_a94cea0ebf2ac4bec69dfa1f80ea07d50"><div class="ttname"><a href="classcutlass_1_1platform_1_1unique__ptr.html#a94cea0ebf2ac4bec69dfa1f80ea07d50">cutlass::platform::unique_ptr::element_type</a></div><div class="ttdeci">T element_type</div><div class="ttdef"><b>Definition:</b> platform.h:715</div></div>
<div class="ttc" id="classcutlass_1_1platform_1_1unique__ptr_html_a5b8d8ecafb4da336acd50e40cd42b6e0"><div class="ttname"><a href="classcutlass_1_1platform_1_1unique__ptr.html#a5b8d8ecafb4da336acd50e40cd42b6e0">cutlass::platform::unique_ptr::get_deleter</a></div><div class="ttdeci">Deleter &amp; get_deleter() noexcept</div><div class="ttdoc">Returns the deleter object. </div><div class="ttdef"><b>Definition:</b> platform.h:757</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1alignment__of_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1alignment__of.html">cutlass::platform::alignment_of</a></div><div class="ttdoc">std::alignment_of </div><div class="ttdef"><b>Definition:</b> platform.h:558</div></div>
<div class="ttc" id="namespacecutlass_1_1platform_html_a57c071d2a7305dd4ec60542e66b0c81c"><div class="ttname"><a href="namespacecutlass_1_1platform.html#a57c071d2a7305dd4ec60542e66b0c81c">cutlass::platform::min</a></div><div class="ttdeci">CUTLASS_HOST_DEVICE constexpr const T &amp; min(const T &amp;a, const T &amp;b)</div><div class="ttdoc">std::min </div><div class="ttdef"><b>Definition:</b> platform.h:183</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1remove__cv_html_a19e5b12cf4eb15ce13d6306735b6de08"><div class="ttname"><a href="structcutlass_1_1platform_1_1remove__cv.html#a19e5b12cf4eb15ce13d6306735b6de08">cutlass::platform::remove_cv::type</a></div><div class="ttdeci">remove_volatile&lt; typename remove_const&lt; T &gt;::type &gt;::type type</div><div class="ttdef"><b>Definition:</b> platform.h:375</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1conditional_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1conditional.html">cutlass::platform::conditional</a></div><div class="ttdoc">std::conditional (true specialization) </div><div class="ttdef"><b>Definition:</b> platform.h:325</div></div>
Updated doxygen
2018-09-20 05:02:08 +08:00
<div class="ttc" id="platform_8h_html_a189faadd7f99f6c354db09acbb2aafcd"><div class="ttname"><a href="platform_8h.html#a189faadd7f99f6c354db09acbb2aafcd">noexcept</a></div><div class="ttdeci">#define noexcept</div><div class="ttdoc">noexcept, constexpr </div><div class="ttdef"><b>Definition:</b> platform.h:134</div></div>
CUTLASS 2.0 (#62) CUTLASS 2.0 Substantially refactored for - Better performance, particularly for native Turing Tensor Cores - Robust and durable templates spanning the design space - Encapsulated functionality embodying modern C++11 programming techniques - Optimized containers and data types for efficient, generic, portable device code Updates to: - Quick start guide - Documentation - Utilities - CUTLASS Profiler Native Turing Tensor Cores - Efficient GEMM kernels targeting Turing Tensor Cores - Mixed-precision floating point, 8-bit integer, 4-bit integer, and binarized operands Coverage of existing CUTLASS functionality: - GEMM kernels targeting CUDA and Tensor Cores in NVIDIA GPUs - Volta Tensor Cores through native mma.sync and through WMMA API - Optimizations such as parallel reductions, threadblock rasterization, and intra-threadblock reductions - Batched GEMM operations - Complex-valued GEMMs Note: this commit and all that follow require a host compiler supporting C++11 or greater.
2019-11-20 08:55:34 +08:00
<div class="ttc" id="classcutlass_1_1platform_1_1unique__ptr_html_a6740f71511f5495d6038cf8878862331"><div class="ttname"><a href="classcutlass_1_1platform_1_1unique__ptr.html#a6740f71511f5495d6038cf8878862331">cutlass::platform::unique_ptr::reset</a></div><div class="ttdeci">void reset(pointer p=pointer()) noexcept</div><div class="ttdoc">Replaces the managed object, deleting the old object. </div><div class="ttdef"><b>Definition:</b> platform.h:745</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1integral__constant_html_a56a164b98a93b1430c26cdfde6f3720d"><div class="ttname"><a href="structcutlass_1_1platform_1_1integral__constant.html#a56a164b98a93b1430c26cdfde6f3720d">cutlass::platform::integral_constant::operator()</a></div><div class="ttdeci">CUTLASS_HOST_DEVICE const value_type operator()() const </div><div class="ttdef"><b>Definition:</b> platform.h:268</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1is__base__of__helper_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1is__base__of__helper.html">cutlass::platform::is_base_of_helper</a></div><div class="ttdoc">Helper for std::is_base_of. </div><div class="ttdef"><b>Definition:</b> platform.h:402</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1remove__const_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1remove__const.html">cutlass::platform::remove_const</a></div><div class="ttdoc">std::remove_const (non-const specialization) </div><div class="ttdef"><b>Definition:</b> platform.h:350</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1aligned__chunk_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1aligned__chunk.html">cutlass::platform::aligned_chunk</a></div><div class="ttdef"><b>Definition:</b> platform.h:631</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1remove__volatile_3_01volatile_01T_01_4_html_aca9bb93efe43106321e4afe0b67542a3"><div class="ttname"><a href="structcutlass_1_1platform_1_1remove__volatile_3_01volatile_01T_01_4.html#aca9bb93efe43106321e4afe0b67542a3">cutlass::platform::remove_volatile&lt; volatile T &gt;::type</a></div><div class="ttdeci">T type</div><div class="ttdef"><b>Definition:</b> platform.h:369</div></div>
<div class="ttc" id="classcutlass_1_1platform_1_1unique__ptr_html_a748d413c50bdbbe9e2f9986fbc423036"><div class="ttname"><a href="classcutlass_1_1platform_1_1unique__ptr.html#a748d413c50bdbbe9e2f9986fbc423036">cutlass::platform::unique_ptr::swap</a></div><div class="ttdeci">void swap(unique_ptr &amp;other) noexcept</div><div class="ttdoc">Swaps the managed objects with *this and another unique_ptr. </div><div class="ttdef"><b>Definition:</b> platform.h:754</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1default__delete_html_a134480d0848e3c213e5a0bee120c90e0"><div class="ttname"><a href="structcutlass_1_1platform_1_1default__delete.html#a134480d0848e3c213e5a0bee120c90e0">cutlass::platform::default_delete::operator()</a></div><div class="ttdeci">void operator()(T *ptr) const </div><div class="ttdef"><b>Definition:</b> platform.h:701</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1aligned__storage_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1aligned__storage.html">cutlass::platform::aligned_storage</a></div><div class="ttdoc">std::aligned_storage </div><div class="ttdef"><b>Definition:</b> platform.h:687</div></div>
<div class="ttc" id="structcutlass_1_1platform_1_1remove__volatile_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1remove__volatile.html">cutlass::platform::remove_volatile</a></div><div class="ttdoc">std::remove_volatile (non-volatile specialization) </div><div class="ttdef"><b>Definition:</b> platform.h:362</div></div>
<div class="ttc" id="classcutlass_1_1platform_1_1unique__ptr_html_a14c8bf5a5deefe4a6602ccd5c5af364c"><div class="ttname"><a href="classcutlass_1_1platform_1_1unique__ptr.html#a14c8bf5a5deefe4a6602ccd5c5af364c">cutlass::platform::unique_ptr::unique_ptr</a></div><div class="ttdeci">unique_ptr(pointer p)</div><div class="ttdef"><b>Definition:</b> platform.h:727</div></div>
<div class="ttc" id="classcutlass_1_1platform_1_1unique__ptr_html_a7ac06ebe7bc66573d3225891e12d2279"><div class="ttname"><a href="classcutlass_1_1platform_1_1unique__ptr.html#a7ac06ebe7bc66573d3225891e12d2279">cutlass::platform::unique_ptr::release</a></div><div class="ttdeci">pointer release() noexcept</div><div class="ttdoc">Releases ownership of the managed object, if any. </div><div class="ttdef"><b>Definition:</b> platform.h:738</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="structcutlass_1_1platform_1_1bool__constant_html"><div class="ttname"><a href="structcutlass_1_1platform_1_1bool__constant.html">cutlass::platform::bool_constant</a></div><div class="ttdoc">std::bool_constant </div><div class="ttdef"><b>Definition:</b> platform.h:288</div></div>
CUTLASS v1.0 release
2018-05-17 02:44:56 +08:00
</div><!-- fragment --></div><!-- contents -->
<!-- start footer part -->
<hr class="footer"/><address class="footer"><small>
CUTLASS 2.0 (#62) CUTLASS 2.0 Substantially refactored for - Better performance, particularly for native Turing Tensor Cores - Robust and durable templates spanning the design space - Encapsulated functionality embodying modern C++11 programming techniques - Optimized containers and data types for efficient, generic, portable device code Updates to: - Quick start guide - Documentation - Utilities - CUTLASS Profiler Native Turing Tensor Cores - Efficient GEMM kernels targeting Turing Tensor Cores - Mixed-precision floating point, 8-bit integer, 4-bit integer, and binarized operands Coverage of existing CUTLASS functionality: - GEMM kernels targeting CUDA and Tensor Cores in NVIDIA GPUs - Volta Tensor Cores through native mma.sync and through WMMA API - Optimizations such as parallel reductions, threadblock rasterization, and intra-threadblock reductions - Batched GEMM operations - Complex-valued GEMMs Note: this commit and all that follow require a host compiler supporting C++11 or greater.
2019-11-20 08:55:34 +08:00
Generated by &#160;<a href="http://www.doxygen.org/index.html">
CUTLASS v1.0 release
2018-05-17 02:44:56 +08:00
<img class="footer" src="doxygen.png" alt="doxygen"/>
CUTLASS 2.0 (#62) CUTLASS 2.0 Substantially refactored for - Better performance, particularly for native Turing Tensor Cores - Robust and durable templates spanning the design space - Encapsulated functionality embodying modern C++11 programming techniques - Optimized containers and data types for efficient, generic, portable device code Updates to: - Quick start guide - Documentation - Utilities - CUTLASS Profiler Native Turing Tensor Cores - Efficient GEMM kernels targeting Turing Tensor Cores - Mixed-precision floating point, 8-bit integer, 4-bit integer, and binarized operands Coverage of existing CUTLASS functionality: - GEMM kernels targeting CUDA and Tensor Cores in NVIDIA GPUs - Volta Tensor Cores through native mma.sync and through WMMA API - Optimizations such as parallel reductions, threadblock rasterization, and intra-threadblock reductions - Batched GEMM operations - Complex-valued GEMMs Note: this commit and all that follow require a host compiler supporting C++11 or greater.
2019-11-20 08:55:34 +08:00
</a> 1.8.11
CUTLASS v1.0 release
2018-05-17 02:44:56 +08:00
</small></address>
</body>
</html>