ml-explore/mlx

[Feature Request] Fallback for custom kernels

Open

#3240 opened on Mar 11, 2026

View on GitHub
 (0 comments) (0 reactions) (0 assignees)C++ (26,216 stars) (1,794 forks)batch import
enhancementgood first issue

Description

We have a specific use-case whereby we need to export MLX callables from Linux. Currently, this prevents us from using any operations that require custom kernels (e.g. grid sample) because otherwise, MLX complains that no Metal GPU was found.

Ideally, we would want to keep the fast kernel so that we can export without a Metal GPU, but execute on a Metal GPU.

Contributor guide

Tech stack
cpp
Domain
machine learningai
Issue type
feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
4
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
over 1 week
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
fresh
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
mostly clear
Prerequisites
C++ knowledgeMLX internalsMetal GPU programming
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
15
Research direction
Research the MLX custom kernel system, specifically how kernels are registered and dispatched. Look at the Metal backend implementation to understand the current limitation. Consider how to create a CPU based fallback using existing array operations or a generic runtime. There are no linked PRs or comments yet.