April 2020

DirectXMath 3.15 includes the following changes:

• Added XMMatrixVectorTensorProduct for creating a matrix from two vectors
• Use of __m256 registers and FMA3 with /arch:AVX2 for stream and some matrix functions
• Optimized load/stores for SSE2 float2 & float3 functions
• Optimized some instruction choices for better AMD CPU support
• Improved conformance for clang/LLVM, GCC, and MinGW compilers
• Code review (constexpr / noexcept usage)
• Retired VS 2015 support

See this post.

This version is also available on NuGet.

Windows 10 May 2020 Update SDK (19041)

DirectXMath 3.14 is included in the Windows 10 SDK (19041)

• Added float control around IsNan functions to resolve issue with VS 2019 with /fp:fast
• XMVerifyCPUSupport updated for clang/LLVM cpuid implementation on x86/x64
• Added support for clang/LLVM built-in platform defines as well as the MSVC ones
• Cleaned up ARM-NEON intrinsics type issues for improved portability
• Removed unneeded malloc.h include in DirectXMath.h
• Whitespace cleanup

See this post

This version is also available on NuGet.

This is the last release to support VS 2015

Windows 10 SDK Preview (18965)

A preview of the DirectXMath 3.14 release is included in the Windows Insider Preview SDK (18945, 18956, 18965).

Note that the AVX2 feature detection with clang/LLVM has a known bug in this version. It's fixed in newer releases.

Windows 10 October 2018 Update SDK (17763)

DirectXMath 3.13 is included in the Windows 10 October 2018 SDK (17763)

• XMFLOAT3X4, XMFLOAT3X4A, and associated Load/Store functions
• Move/copy constructors and assignment operators for C++ types
• Minor fix for XMVectorClamp behavior with NaN
• Fixed compilation warnings with VS 2017 (15.7 update), Intel C++ 18.0 compiler, and clang 6
• Retired VS 2013 support
• Minor code cleanup

See this post

This version is also available on NuGet.

VS 2017 (15.9 update) supports Universal Windows Platform (UWP) apps for the ARM64 (ARMv8) architecture which is supported by DirectXMath starting with version 3.07.

Windows 10 SDK Preview (17713)

This is a preview release of DirectXMath 3.13 from the Windows 10 SDK Insider Preview (17713).

Windows 10 SDK Preview (17709)

This is a preview release of DirectXMath 3.13 from the Windows 10 SDK Insider Preview (17704, 17709).

Windows 10 SDK Preview (17692)

This is a preview release of DirectXMath 3.13 from the Windows 10 SDK Insider Preview (17677, 17682, 17686, 17692).

Windows 10 SDK Preview (17672)

This is a preview release of DirectXMath 3.13 from the Windows 10 SDK Insider Preview (17666, 17672).

February 2018

DirectXMath 3.12 is included in the Windows Insider Preview SDK (17661)

• ARM64 use of fused multiply-accumulate intriniscs
• Conformance fix for XMConvertFloatToHalf
• Minor code cleanup

This version is also available on NuGet.

Last release to support VS 2013, Windows Store 8.1, or Windows Phone 8.1

Windows 10 Fall Creators Update SDK (16299)

DirectXMath 3.11 is included in the Windows 10 Fall Creators Update SDK (16299) as well as the Windows 10 April 2018 Update SDK (17134).

• AVX optimization of XMMatrixMultiply and XMMatrixMultiplyTranspose
• AVX2 optimization for XMVectorSplatX
• FMA3 optimization of XMVectorMultiplyAdd and XMVectorNegativeMultiplySubtract (implied by /arch:AVX2)
• Conformance fixes to support compilation with Clang 3.7

See this post.

This version is also available on NuGet.