AuroraMiddleware/DirectXMath
1
0
Fork 0
mirror of https://github.com/microsoft/DirectXMath synced 2024-11-23 12:50:05 +00:00
Code Issues Releases Wiki Activity

Updated Home (markdown)

Chuck Walbourn 2021-07-10 13:48:56 -07:00
parent 3aa0c43868
commit 385ed48a91
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
Home.md

@ -118,7 +118,7 @@ For non-Windows platforms, you should use the open source version of [sal.h](htt
## Intel Short Vector Math Library
In Visual Studio 2019, the Intel® Short Vector Library is now available as part of the Visual C++ runtime libraries. In DirectXMath 3.16 or later, I make use of this library to replace the minimax approximations of various transcendental and other math functions listed below. With the GitHub, NuGet, and VCPKG versions this is enabled automatically, but can be disabled by defining the preprocessor symbol `_XM_DISABLE_INTEL_SVML_`. With the Windows 11 SDK, developers must **opt-in** by defining the preprocessor symbol `_XM_SVML_INTRINSICS_`.
In Visual Studio 2019, the Intel® Short Vector Library is [now available](https://devblogs.microsoft.com/cppblog/msvc-backend-updates-in-visual-studio-2019-preview-2/) as part of the Visual C++ runtime libraries. In DirectXMath 3.16 or later, I make use of this library to replace the minimax approximations of various transcendental and other math functions listed below. With the GitHub, NuGet, and VCPKG versions this is enabled automatically, but can be disabled by defining the preprocessor symbol `_XM_DISABLE_INTEL_SVML_`. With the Windows 11 SDK, developers must **opt-in** by defining the preprocessor symbol `_XM_SVML_INTRINSICS_`.
For the x86/x64 platform, enabling the SVML makes the following DirectXMath functions into simple inline wrappers around the appropriate SVML function:
@ -134,7 +134,7 @@ For the x86/x64 platform, enabling the SVML makes the following DirectXMath func
<tr><td>XMVectorATan<br />XMVectorATanEst</td><td>_mm_atan_ps</td><td></td><td>XMVectorATan2<br />XMVectorATan2Est</td><td>_mm_atan2_ps</td></tr>
</table>
> The Intel&reg; Short Vector Library routines are generally optimized for SSE/SSE2, SSE 4, and AVX selected at runtime. They are work well on both Intel and AMD CPUs. See the [Intel site](https://software.intel.com/content/www/us/en/develop/documentation/cpp-compiler-developer-guide-and-reference/top/compiler-reference/intrinsics/intrinsics-for-short-vector-math-library-operations/) for more information.
# Additional content