Who wants a revamped Mac Professional when you possibly can simply flip a number of Mac Studios right into a unified computing system? With the upcoming macOS Tahoe 26.2 launch, Apple is introducing a brand new low-latency characteristic that permits you to join a number of Macs collectively utilizing Thunderbolt 5. For builders and researchers, it is a doubtlessly helpful solution to create highly effective AI supercomputers that may run large native fashions. That enables 4 Mac Studios, which might every run as much as 512GB of unified reminiscence, to run the 1 trillion parameter Kimi-K2-Considering mannequin way more effectively than PCs with power-hungry GPUs.

Whereas we’ve seen Thunderbolt Mac clusters before, they have been restricted by slower Thunderbolt speeds, particularly in the event that they required a hub (which may cut back speeds to 10 Gb/s). Apple’s new characteristic permits for the complete Thunderbolt 5 connectivity of as much as 80Gb/s. The clustering functionality additionally is not simply restricted to the dear Mac Studio, it’ll additionally work with the M4 Professional Mac mini and M4 Professional/Max MacBook Professional. Builders will not want any particular {hardware} to construct clusters, simply normal Thunderbolt 5 cables and appropriate Macs.

In a demo, I watched as a cluster of 4 Mac Studios loaded and ran that large Kimi-K2-Considering mannequin in an early model of ExoLabs’s EXO 1.0. Notably, the cluster used lower than 500 watts of energy, which is round 10 instances decrease than a typical GPU cluster (NVIDIA’s RTX 5090 is rated for 575W, however its calls for also can leap greater).

macOS Tahoe 26.2 can even give Apple’s open supply MLX venture full entry to the neural accelerators on the M5 chip, which ought to dramatically pace up AI inferencing. Paradoxically, although, the one M5 Mac out there immediately — the 14-inch MacBook Professional — solely helps Thunderbolt 4. Which means it received’t have the ability to make the most of the brand new Mac clustering functionality.

The unified reminiscence and low energy design of Apple Silicon already made Macs a helpful selection for demanding AI work, however the capacity to cluster a number of programs collectively over Thunderbolt 5 is doubtlessly much more tempting to anybody working with massive fashions. After all, a Mac Studio with 512GB of RAM is not low-cost — it begins at $9,499 with the M3 Extremely chip — however that is solely the highest-end choice. Labs and firms that have already got Mac Studios, Mac minis and MacBook Execs may doubtlessly cluster programs they’ve already bought.