Apple Unveils M6 and M5 Ultra: 2nm Silicon Built for On-Device AI
Apple introduced its first 2-nanometer chip M6 in a new Mac mini and the quad-die M5 Ultra in a new Mac Studio, with up to 512GB of unified memory and 1.2TB/s bandwidth. What it means for local AI.
What happened
On August 25, 2026, Apple introduced the M6 — its first 2-nanometer chip — in a new Mac mini, alongside the M5 Ultra in a new Mac Studio. The M5 Ultra is Apple's most powerful chip ever and the company's first quad-die system on a chip.
M6: 2nm and a Dual Neural Engine
M6 packs a larger 12-core CPU with the world's fastest CPU core, a 12-core GPU with a Neural Accelerator in each core, and a Dual 16-core Neural Engine that roughly doubles peak compute for on-device AI. Memory bandwidth reaches up to 170GB/s.
M5 Ultra: quad-die power
M5 Ultra joins two M5 Max dies through next-generation UltraFusion — a first for Apple silicon. It offers up to a 36-core CPU, up to an 80-core GPU, 512GB of unified memory and 1.2TB/s of bandwidth, 50 percent more than M3 Ultra.
Why it matters
Both chips work with Core AI, Core ML, Metal and Xcode, so developers can run and fine-tune large models locally. The message to enterprises: frontier-class AI without the cloud.