GoKawiilApple today announced its new M6 processor, launching first on the popular Mac Mini. It's been clear for several years that Apple pulled off a rare coup when it switched away from Intel to its own, ARM-based silicon, so the real question has been the degree to which those initial efficiency gains would diminish generation-over-generation. How does the new M6 stack up against its predecessor, the M5?
With the base M6, Apple is finally shipping a chip built on a 2nm process, a leap forward Mac users have been crossing their fingers for. The new chip also gets a bump from 10 to 12 CPU cores (two super cores, four performance cores and six efficiency cores), and similarly receives a 12-core GPU alongside a dual 16-core NPU. The company is claiming that the M6 has "the world's fastest CPU core" and boasts faster memory bandwidth, an accelerated GPU and up to 170GB/s of unified memory bandwidth. Even so, Cupertino is clear that the M6 is intended for everyday use cases like school, work and hobbyist AI.
Compared to the last-generation M5 SoC, the M6 is a reasonable on-paper upgrade, with Apple asserting a one-fifth increase in performance for multithreaded tasks, and an even bigger uplift in GPU-bound performance for AI tasks. We'll need thorough, hands-on testing to confirm those claims, so stay tuned. It will also have a base configuration of 16GB on the refreshed Mac Mini, or up to 32GB, though you'll pay for the privilege with Apple finally passing along to consumers the costs of the RAM crisis driven by AI demand. Here's how the refreshed M6 silicon shakes out compared to the M5.