Apple rocked the computing world with its M1 chip, the first "Apple Silicon" hardware that turned the MacBook Air, Mac Mini and other computers into portable powerhouses. Last year, the company followed that up with the M1 Pro and M1 Max, which delivered even more performance for the 14-inch and 16-inch MacBook Pro. Now, Apple is adding a new member to the family: the M1 Ultra.
The M1 Ultra is essentially two M1 Max chips put together, making it even better suited to intensive creative applications like video editing and 3D rendering. During its launch event today, Apple revealed that the M1 Max chips housed a secret feature: a die-to-die interconnect, dubbed "UltraFusion," that allows it to connect multiple chips. Conceptually, it's similar to AMD's Infinity Fabric, which ensures speedy communication between the CPU, GPU and other components.
Apple says the UltraFusion interconnect can handle bandwidth up to 2.5 terabytes per second, so it shouldn't lead to any performance slowdowns between the two M1 Max dies. Altogether, the M1 Ultra sports a whopping 114 billion transistors, and it supports up to 128GB of unified memory with 800 GB/s of bandwidth. As you'd expect, its specs are basically what happens when you sandwich two M1 Max chips: the Ultra features a 20-core CPU (16 high performance and 4 high efficiency cores), and a 64-core GPU. The company claims it offers up to 8 times faster graphics than the original M1 chip.
Given that the M1 Ultra will make its debut in Apple's new Mac Studio mini-desktop, the company didn't need to worry about battery life at all. Still, Apple says the Ultra is at least more efficient than the competition, as it uses up to 65 percent less power than a 10-core x86 chip. Naturally, Apple didn't reveal which CPU it was comparing the M1 Ultra to, but the numbers make sense given what we've seen from the M1 Max so far.
Catch up on all of the news from Apple’s Peek Performance event right here!
