Implementations of SIMD instruction sets for systems which don't natively support them.

Expressive Vector Engine - SIMD in C++ Goes Brrrr

Portable header-only C++ low level SIMD library

A header only library implementing common mathematical functions using SIMD intrinsics

UME::SIMD A library for explicit simd vectorization.

SIMD macro assembler unified for ARM, MIPS, PPC and x86

Realtime raytracer using SIMD on ARM, MIPS, PPC and x86

Non-bijunctive attention collapse for LLM inference — POWER8 hardware AES (vcipher) + AltiVec vec_perm. Hebbian path selection, cross-head diffusion, O(1) KV prefiltering.

ArchLoader is a utility to unlock the Halo 4 Network Test and add extensive functionality otherwise not available (even when the beta was active).

MacGLide is a shared library for Apple Macintosh computers that emulates 3Dfx Voodoo Graphics hardware. It makes it possible to run games originally made for 3Dfx/Glide without the hardware by translating Glide-function calls to OpenGL.

Local libvpx changes (POWER8 Altivec/VSX support)

Mandelbrot and Julia fractals for AmigaOS4 with Altivec vector unit support

PowerPC Altivec to C plugin for IDA

Accelerate LLM inference by collapsing attention paths with hardware-optimized selective pruning using POWER8 vector instructions and crypto operators.

Local branches for https://aomedia.googlesource.com/aom

Experiments with adding SIMD support to Rust's standard library.

2-norm guided FP32 truncation for heterogeneous deep learning training