v0.40.7-rc1: mlx: deduplicate CUDA runtime payloads (#17956)
Deduplicate MLX CUDA runtime payloads Move CUDA libraries shared by the GGML and MLX backends into the sibling CUDA payload, and teach in-process and subprocess MLX…
What moved
Deduplicate MLX CUDA runtime payloads Move CUDA libraries shared by the GGML and MLX backends into the sibling CUDA payload, and teach in-process and subprocess MLX loading to resolve that packaged dependency directory. …
Why it matters
Filed as a launch filing from Ollama, 2026-10-20. The desk tags this brief Hardware.
On the record
- Filed from the Ollama official GitHub releases feed on 2026-10-20.
- Primary source host: github.com.
- Deduplicate MLX CUDA runtime payloads Move CUDA libraries shared by the GGML and MLX backends into the sibling CUDA payload, and teach in-process and subprocess MLX loading to resolve that packaged dependency directory.
Primary source
github.com
github.com
Desk
Logged as brief 151
Logged as brief 151
Cite this brief
v0.40.7-rc1: mlx: deduplicate CUDA runtime payloads (#17956) (Ollama, 2026-10-20) — https://labledgerdesk.pages.dev/b/2026/10/20/v0-40-7-rc1-mlx-deduplicate-cuda-runtime-payloads-17956/