Chuck Walbourn

Per the [porting guide](https://www.intel.com/content/www/us/en/developer/articles/guide/porting-guide-for-icc-users-to-dpcpp-or-icx.html?msclkid=d1726d45d15f11eca19c0a78074dff09), there the control define for the new Intel LLVM compiler is ``__INTEL_LLVM_COMPILER``. Should update DirectXMath to support both the new and classic Intel compilers.

Add XMMatrixInverseTranspose

1

The DirectXMath library includes both a `XMMatrixTranspose` and a `XMMatrixInverse`, but doing both at once could be made more efficient by omitting the initial transpose.

Add XMVectorRound half away from zero alternative

2

The current `XMVectorRound` uses round-to-nearest (even) a.k.a. _banker's rounding_. This matches the implementation of the `_mm_round_ps` (SSE4) and `vrndnq_f32` (ARMv8 NEON) intrinsics rounding behavior, so it can be implemented in...

The original D3DX9math implementation of Matrix Stack only tracked a single matrix. One of the reasons, however, to create a stack is to be able to step-by-step maintain a matching...

BoundingOrientedBox::CreateFromPoints solver suffers from numeric problems

2

Report from Dave Eberly about problems with the oriented box `SolveCubic` helper function. > The SolveCubic is called to compute the eigenvalues of a real-valued symmetric 3x3 matrix. Such matrices...

Revisit XMVector*Orthogonal

1

See [Building an Orthonormal Basis from a 3D Unit Vector Without Normalization ](http://orbit.dtu.dk/files/126824972/onb_frisvad_jgt2012_v2.pdf) and [Building an Orthonormal Basis, Revisited ](https://graphics.pixar.com/library/OrthonormalB/paper.pdf)

In the original Xbox 360 implementation of xboxmath, the following functions used tricks and helper intrinsics like ``vec_ctf`` to generate constants: * ``XMVectorSplatOne`` * ``XMVectorSplatInfinity`` * ``XMVectorSplatQNaN`` * ``XMVectorSplatEpsilon`` *...

The `XMLoadFloat3SE` and `XMStoreFloat3SE` functions are both all scalar. They could benefit from SIMD optimization. > Note this matches `DXGI_FORMAT_R9G9B9E5_SHAREDEXP`

The `XMLoadFloat3PK` and `XMStoreFloat3PK` functions are both all scalar. They could benefit from SIMD optimization. > This format matches `DXGI_FORMAT_R11G11B10_FLOAT`

DirectXMath has a number of different forms of transformation, but currently lacks an optimized function for doing transforms by transposed matrices: - `XMVector3TransformTranspose` - `XMVector4TransformTranspose` - `XMVector3TransformTransposeStream` - `XMVector4TransformTransposeStream` This...