Apple switched to its own computer chips three years ago, boldly moving toward complete control of its technology stack. today, Apple launched MLXan open source framework specifically designed to perform machine learning on Apple M CPUs.
Most AI software is currently developed on Linux or Microsoft’s open source systems, and Apple doesn’t want to exclude its thriving developer ecosystem from the latest systems.
MLX aims to solve long-standing compatibility and performance issues associated with Apple’s unique architecture and software, but it’s more than just a technology toy. MLX offers a user-friendly design, perhaps inspired by well-known frameworks such as PyTorch, Jax, and ArrayFire. Its introduction promises a more efficient process for training and deploying AI machine learning models on Apple devices.
Architecturally, MLX features a unified memory model, where arrays reside in shared memory, allowing operations on different types of supported devices without the need for data duplication. This feature is essential for developers looking for flexibility in their AI projects.
In short, unified memory means that your GPU shares its VRAM with your computer’s RAM, so instead of buying a powerful PC and then adding a powerful GPU that has a lot of vRAM, you can simply use your Mac’s RAM For everything.
However, the path to AI development on Apple Silicon has not been without challenges, mainly due to its closed ecosystem and lack of support for many open source development projects and widely used infrastructure.
“It’s exciting to see more tools like this for working with string-like objects, but I really wish Apple would make it easier to port custom models more efficiently in terms of performance,” one developer said. In Hacker News In a discussion about advertising.
Until now, developers had to convert their models to CoreML in order to be able to run them on Apple. This depends on the translator It’s not perfect.
CoreML focuses on transforming pre-existing machine learning models and optimizing them for Apple devices. While MLX is about building machine learning models and running them directly and efficiently on Apple’s own devices, providing tools for innovation and development within the Apple ecosystem.
The MLX performed well in benchmark tests. Its compatibility with tools like Stable Diffusion and OpenAI’s Whisper represents a major advance. Most importantly, performance comparisons reveal MLX’s efficiency, outperforming PyTorch in imaging speeds at larger batch sizes.
For example, Apple Tells It takes “about 90 seconds to fully generate 16 images using MLX, 50 publishing steps with classifier-free directives, and about 120 using PyTorch.”
As artificial intelligence continues to evolve at a rapid pace, MLX represents an important milestone for the Apple ecosystem. It not only addresses technical challenges, but also opens up new possibilities for AI and machine learning R&D on Apple devices, a strategic move given Apple’s separation from Nvidia and its robust AI ecosystem.
MLX aims to make Apple’s platform a more attractive and feasible option for AI researchers and developers, and means a brighter Christmas for Apple’s AI-obsessed fans.
“Beer enthusiast. Subtly charming alcohol junkie. Wannabe internet buff. Typical pop culture lover.”