63 lines
7.2 KiB
HTML
63 lines
7.2 KiB
HTML
<html>
|
|
<head>
|
|
<link rel="stylesheet" type="text/css" href="res/style.css">
|
|
<link rel="icon" href="res/logo.ico" type="image/x-icon"/>
|
|
<meta name="viewport" content="width=device-width, initial-scale=1.0, maximum-scale=1.0, minimum-scale=1.0">
|
|
<title>NVIDIA
|
|
Nsight Perf SDK Release Notes - v2021.1 Early Access 2
|
|
</title>
|
|
</head>
|
|
|
|
<body>
|
|
<img src="res/logo.png" width="100"/>
|
|
<h1>Nsight Perf SDK Release Notes - v2021.1 Early Access 2</h1>
|
|
<h2>Table of Contents</h2>
|
|
<ul class="toc">
|
|
<li><a href="#prerequisites">Prerequisites</a></li>
|
|
<li><a href="#known_issues">Known Issues</a></li>
|
|
<li><a href="#limitations">Limitations</a></li>
|
|
<li><a href="#0xe40360ec102bde77">v2021.1 Early Access 2</a></li>
|
|
<li><a href="#0xf0d94cae676d012a">v2021.1 Early Access 1</a></li>
|
|
</ul>
|
|
|
|
<h2><a name="prerequisites">Prerequisites</a></h2>
|
|
<ul class="prerequisites">
|
|
<li>Windows 10 or Linux (Ubuntu 20.04 or similar distro).</li>
|
|
<li>NVIDIA Graphics Driver: R465 or newer.</li>
|
|
<li>NVIDIA GeForce, Quadro, or RTX GPU listed below:<br>
|
|
<table class='chips'><thead>
|
|
<tr><th>NVIDIA Architecture</th> <th>Chip Names</th> <th>GeForce Families</th> <th>Quadro and NVIDIA RTX Families</th></tr>
|
|
<tr><th>NVIDIA Ampere GA10x Architecture</th> <th>GA102, GA104, GA106</th> <th>RTX 3090, RTX 3080, RTX 3070, RTX 3060</th> <th>RTX A6000, A40</th></tr>
|
|
<tr><th>NVIDIA Turing TU10x Architecture</th> <th>TU102, TU104, TU106</th> <th>Titan RTX, RTX 2080, RTX 2070, RTX 2060</th> <th>RTX 8000, RTX 6000, RTX 5000, RTX 4000, RTX 3000</th></tr>
|
|
<tr><th>NVIDIA Turing TU11x Architecture</th> <th>TU116, TU117</th> <th>GTX 1660, GTX 1650</th> <th>T2000, T1000</th></tr>
|
|
<tr><th>NVIDIA Volta V100i Architecture</th> <th>GV100</th> <th>Titan V</th> <th>Quadro GV100</th></tr>
|
|
</thead></table></li>
|
|
<li>By default, the driver disallows GPU profiling. To enable permissions, follow the instructions at <a href="https://developer.nvidia.com/ERR_NVGPUCTRPERM">https://developer.nvidia.com/ERR_NVGPUCTRPERM</a>, or try the workarounds described in the Getting Started Guide.</li>
|
|
</ul>
|
|
<h2><a name="known_issues">Known Issues</a></h2>
|
|
<ul class="known_issues">
|
|
<li>vkQueueWaitIdle is required before present when profiling Vulkan applications. (<span class="gray">3201058</span>)</li><li>Vulkan profiling on desktop linux requires X-windows to be running. (<span class="gray">3289536</span>)</li></ul>
|
|
<h2><a name="limitations">Limitations</a></h2>
|
|
<ul class="limitations">
|
|
<li>Nsight Perf SDK does not support SLI systems. (<span class="gray">PW-2935</span>)</li><li>Each GPU is limited to one profiling session at a time, across the entire system. However, multiple GPUs can be independently profiled concurrently. This is a driver limitation.</li><li>D3D12 Range Profiler is not compatible with the D3D12 debug layer. (<span class="gray">PW-925</span>)</li><li>D3D12 Range Profiler does not support PushRange and PopRange commands on Bundles (Command Lists created with <a href="https://docs.microsoft.com/en-us/windows/win32/api/d3d12/ne-d3d12-d3d12_command_list_type">D3D12_COMMAND_LIST_TYPE_BUNDLE</a>).</li><li>D3D11 Range Profiler does not support PushRange and PopRange calls on a <a href="https://docs.microsoft.com/en-us/windows/win32/direct3d11/overviews-direct3d-11-render-multi-thread-render">Deffered Context</a>.</li><li>Vulkan Range Profiler is not compatible with Vulkan layers that perform <a href="https://vulkan.lunarg.com/doc/view/1.1.114.0/linux/loader_and_layer_interface.html#user-content-wrapping">object-wrapping</a>.</li><li>Vulkan Range Profiler does not support PushRange and PopRange calls on secondary command buffers (VkCommandBuffer created with <a href="https://www.khronos.org/registry/vulkan/specs/1.2-extensions/man/html/VkCommandBufferLevel.html">VK_COMMAND_BUFFER_LEVEL_SECONDARY</a>.</li><li>Vulkan Range Profiler does not support PushRange and PopRange calls on a VkCommandBuffer that is recording with <a href="https://www.khronos.org/registry/vulkan/specs/1.2-extensions/man/html/VkCommandBufferUsageFlagBits.html">VK_COMMAND_BUFFER_USAGE_SIMULTANEOUS_USE_BIT</a>.</li><li>NVPW_Device_SetClockSetting() performs a system-global clock change, that is not undone at process exit. The *SetDeviceClockState() family of NvPerfUtility functions have the same behavior, as they are convenience functions around NVPW_Device_SetClockSetting().</li><li>Not all counters and metrics are available to the Range Profiler APIs, due to driver limitations. The CounterAvailability APIs exist to reveal which metrics are available.</li></ul>
|
|
<h2><a name="0xe40360ec102bde77">v2021.1 Early Access 2</a></h2>
|
|
<ul class="changelog">
|
|
<li>API Versions<br>
|
|
<table class="api">
|
|
<thead>
|
|
<tr>
|
|
<th>API</th><th>New Version</th><th>Previous Version</th><tr><td>GV100 Graphics Metrics</td><td style="text-align:center">0.1.0</td><td style="text-align:center">0.1.0</td></tr><tr><td>TU10X Graphics Metrics</td><td style="text-align:center">0.1.0</td><td style="text-align:center">0.1.0</td></tr><tr><td>TU11X Graphics Metrics</td><td style="text-align:center">0.1.0</td><td style="text-align:center">0.1.0</td></tr><tr><td>GA10X Graphics Metrics</td><td style="text-align:center">0.1.0</td><td style="text-align:center">0.1.0</td></tr><tr><td>Metrics Evaluator</td><td style="text-align:center">0.1.0</td><td style="text-align:center">0.1.0</td></tr><tr><td>D3D11 Profiler</td><td style="text-align:center">0.1.0</td><td style="text-align:center">0.1.0</td></tr><tr><td>D3D12 Profiler</td><td style="text-align:center">0.1.0</td><td style="text-align:center">0.1.0</td></tr><tr><td>OpenGL Profiler</td><td style="text-align:center">0.1.0</td><td style="text-align:center">0.1.0</td></tr><tr><td>Vulkan Profiler</td><td style="text-align:center">0.1.0</td><td style="text-align:center">0.1.0</td></tr></tbody>
|
|
</table></li>
|
|
</ul>
|
|
<h2><a name="0xf0d94cae676d012a">v2021.1 Early Access 1</a></h2>
|
|
<ul class="changelog">
|
|
<li>API Versions<br>
|
|
<table class="api">
|
|
<thead>
|
|
<tr>
|
|
<th>API</th><th>New Version</th><th>Previous Version</th><tr><td>GV100 Graphics Metrics</td><td style="text-align:center">0.1.0</td><td style="text-align:center">N/A</td></tr><tr><td>TU10X Graphics Metrics</td><td style="text-align:center">0.1.0</td><td style="text-align:center">N/A</td></tr><tr><td>TU11X Graphics Metrics</td><td style="text-align:center">0.1.0</td><td style="text-align:center">N/A</td></tr><tr><td>GA10X Graphics Metrics</td><td style="text-align:center">0.1.0</td><td style="text-align:center">N/A</td></tr><tr><td>Metrics Evaluator</td><td style="text-align:center">0.1.0</td><td style="text-align:center">N/A</td></tr><tr><td>D3D11 Profiler</td><td style="text-align:center">0.1.0</td><td style="text-align:center">N/A</td></tr><tr><td>D3D12 Profiler</td><td style="text-align:center">0.1.0</td><td style="text-align:center">N/A</td></tr><tr><td>OpenGL Profiler</td><td style="text-align:center">0.1.0</td><td style="text-align:center">N/A</td></tr><tr><td>Vulkan Profiler</td><td style="text-align:center">0.1.0</td><td style="text-align:center">N/A</td></tr></tbody>
|
|
</table></li>
|
|
</ul>
|
|
|
|
</body>
|
|
</html> |