You can turn a cluster of Macs into an AI supercomputer in macOS Tahoe 26.2
Who needs a revamped Mac Pro when you can just turn several Mac Studios into a unified computing system? With the upcoming macOS Tahoe 26.2 release, Apple is introducing a new low-latency feature that lets you connect several Macs together using Thunderbolt 5. For developers and researchers, it's a potentially useful way to create powerful AI supercomputers that can run massive local models. That allows four Mac Studios, which can each run up to 512GB of unified memory, to run the 1 trillion parameter Kimi-K2-Thinking model far more efficiently than PCs with power-hungry GPUs.While we’ve seen Thunderbolt Mac clusters before, they were limited by slower Thunderbolt speeds, especially if they required a hub (which could reduce speeds to 10 Gb/s). Apple’s new feature allows for the full Thunderbolt 5 connectivity of up to 80Gb/s. The clustering capability also isn't just limited to the pricey Mac Studio, it will also work with the M4 Pro Mac mini and M4 Pro/Max MacBook Pro. Developers won't need any special hardware to build clusters, just standard Thunderbolt 5 cables and compatible Macs. In a demo, I watched as a cluster of four Mac Studios loaded and ran that massive Kimi-K2-Thinking model in an early version of ExoLabs's EXO 1.0. Notably, the cluster used less than 500 watts of power, which is around 10 times lower than a typical GPU cluster (NVIDIA’s RTX 5090 is rated for 575W, but its demands can also jump higher).macOS Tahoe 26.2 will also give Apple’s open source MLX project full access to the neural accelerators on the M5 chip, which should dramatically speed up AI inferencing. Ironically, though, the only M5 Mac available today — the 14-inch MacBook Pro — only supports Thunderbolt 4. That means it won’t be able to take advantage of the new Mac clustering capability. The unified memory and low power design of Apple Silicon already made Macs a useful choice for demanding AI work, but the ability to cluster multiple systems together over Thunderbolt 5 is potentially even more tempting to anyone working with large models. Of course, a Mac Studio with 512GB of RAM isn't cheap -- it starts at $9,499 with the M3 Ultra chip -- but that's only the highest-end option. Labs and companies that already have Mac Studios, Mac minis and MacBook Pros could potentially cluster systems they've already purchased.This article originally appeared on Engadget at https://www.engadget.com/ai/you-can-turn-a-cluster-of-macs-into-an-ai-supercomputer-in-macos-tahoe-262-191500778.html?src=rss