New capability:

ggerganov/llama.cpp b9222: b9222

GitHub·AI·NewCapability·5/19/2026

<details open> hexagon: add support for TRI op (#22822) * Hexagon: TRI HVX Kernel addition to ggml hexagon HTP ops and context * addressed PR review comments for TRI op * hexagon: clang format * hex-unary: remove merge conflict markers * hex-ggml: remove duplicate op cases (merge conflict) *

Why it matters → New functionality is available in llama.cpp. May enable simpler implementations or replace external dependencies.

Who should care → Teams already using llama.cpp or evaluating it.

llama.cpp

View original source ↗

Source payload preview

{
  "id": 324650308,
  "url": "https://api.github.com/repos/ggml-org/llama.cpp/releases/324650308",
  "body": "<details open>\n\nhexagon: add support for TRI op (#22822)\n\n* Hexagon: TRI HVX Kernel addition to ggml hexagon HTP ops and context\n\n* addressed PR review comments for TRI op\n\n* hexagon: clang format\n\n* hex-unary: remove merge conflict markers\n\n* hex-ggml: remove duplicate op cases (merge conflict)\n\n* hex-ggml: fix editor config errors\n\n---------\n\nCo-authored-by: Todor Boinovski <todorb@qti.qualcomm.com>\nCo-authored-by: Max Krasnyansky <maxk@qti.qualcomm.com>\n\n</details>\n\n**macOS/iOS:**\n- [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b9222/llama-b9222-bin-macos-arm64.tar.gz)\n- [macOS Apple Silicon (arm64, KleidiAI enabl
…