Every M-series launch I do the same math and land the same place. M6 gets a Dual 16-core Neural Engine, 170GB/s of bandwidth, Neural Accelerators in all twelve GPU cores — and 32GB of unified memory, which is the number that decides what you can actually run. Faster prompt processing is real, but it doesn’t change which model fits.
So my plan stays split. The mini stays the always-on box for small local work: indexing, compiling, agentic tasks. Anything with hundreds of billions of parameters wants the 512GB M5 Ultra, and that’s a Mac Studio purchase. Apple is honest about the split — mini for AI hobbyists, Ultra for frontier models on device.
The story — Apple debuted M6 in the new Mac mini and M5 Ultra in the new Mac Studio. M6 is Apple’s first 2nm chip, with a 12-core CPU, 12-core GPU, a Dual 16-core Neural Engine, up to 32GB of unified memory and 170GB/s of bandwidth. M5 Ultra uses UltraFusion to join two dual-die M5 Max chips into Apple’s first quad-die SoC: up to 36-core CPU, 80-core GPU, 512GB memory, 1.2TB/s bandwidth. (Source)