-
|
Hello and thanks for the great work on the MLX, I enjoy writing research code in it, but I'm a bit hesitant to share it as some of my colleagues / friends won't be able to use it due to the ARM dependency. Are there any plans for some sort of fallback simulation mode that would run on intel processors / other platforms? |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 1 reply
-
|
We actually have a generic back-end that can run on Intel CPUs. Right now it has to be built from source, but we could make a pypi wheel for it. It's not a big priority for us so it may not be as well optimized as getting something platform specific. We do not have plans to do a CUDA backend or other accelerators in the near future. |
Beta Was this translation helpful? Give feedback.
-
|
Excellent, that would be sufficient I think. Understandable re: CUDA backend, etc. |
Beta Was this translation helpful? Give feedback.
We actually have a generic back-end that can run on Intel CPUs. Right now it has to be built from source, but we could make a pypi wheel for it. It's not a big priority for us so it may not be as well optimized as getting something platform specific.
We do not have plans to do a CUDA backend or other accelerators in the near future.