Developer-Ecosystem-Engineering

Mac-videotoolbox: Add support for VTB ProRes and HEVC encoding

7

### Description Adds support for additional hardware encoders for HEVC and ProRes and prefers to use hardware encoders when they are available **New Encoder List** **New Encoder Options** **New Profile...

Buffers in pool may not always point at the video compositors output

2

We received a report of black frames in an applications video content once it was encoded with NextLevelSessionExporter. Investigation revealed a similar issue as https://github.com/NextLevel/NextLevelSessionExporter/issues/38. It appears NextLevelSessionExporter is pulling...

Use NEON when running on arm64 Add SIMD test Tested on macOS 11.2 + M1 Mac Mini Running ./main before patch with subset_size_milllions = 10 Parsing gt: 10000 Loaded gt...

Workloads not properly configured for `aarch64` and Apple silicon

2

Hi Phoronix, We noticed issues with the Phoronix test suite where the results measured may not be reflective of actual machine capabilities. This may impact other `aarch64` configurations beyond Apple...

This patch avoids behaviors that are not performant on Apple silicon: spin waiting and yielding. This also replaces a spin lock with os_unfair_lock as that’s a better lock for macOS....

NumPy has SIMD versions of BOOL `logical_and`, `logical_or`, `logical_not`, and `absolute` for SSE2. The changes here replace that implementation with one that uses universal intrinsics. This allows other architectures to...

ENH: Implement SIMD versions of isnan,isinf, isfinite and signbit

6

NumPy has SIMD versions of float / double `isnan`, `isinf`, `isfinite`, and `signbit` for SSE2 and AVX-512. The changes here replace the SSE2 version with one that uses universal intrinsics....

Optimize NEON encoding for high dynamic range blocks

6

We can improve the encoding performance on Apple silicon configurations with the following enhancements - Replace the switch with a for loop - Invert qlp_factors & remove else conditions as...

Read/Write dependencies on constraints in contact solvers impacting out of order processors

11

The [contact solver ](https://github.com/erincatto/box2d/blob/c6cc3646d1701ab3c0750ef397d2d68fc6dbcff2/src/dynamics/b2_contact_solver.cpp#L51) creates constraints between Object A and Object B. When processing these we read values from A and B, and write results to A/B. However, constraints can...

Updating to the latest Catch2 will allow `expected` to also compile for Apple silicon natively. Once updated % mkdir build % cd build % cmake .. % make % make...