The First 2nm Apple Silicon Is Finally on Shelves
Apple introduced the M6 in its new Mac mini on August 25, marking the company's first chip built on TSMC's 2-nanometer process. The same launch brought the M5 Ultra, Apple's most powerful chip ever, inside a redesigned Mac Studio. Together they close out the long-rumored M6 transition and reset the Mac's performance and AI ambitions.
The 2nm node ends a run of three consecutive 3nm generations spanning the M3, M4, and M5. Packing more transistors onto a smaller die buys both higher clocks and better power efficiency, the two levers that matter most in a desktop that now starts at $899.
What the 2nm M6 Actually Delivers
The M6 pairs a larger 12-core CPU with a 12-core GPU and a Dual 16-core Neural Engine. Apple calls its new super cores the world's fastest single-threaded CPU cores, and the full array runs about 1.2x faster than the M5 in multithreaded work and up to 2.4x faster than the M1.
Every GPU core now carries its own Neural Accelerator, lifting peak GPU compute for AI by roughly 30 percent over the M5 and more than 8x over the M1. The Dual 16-core Neural Engine doubles peak compute versus earlier generations, and apps can use both engines at once to speed up on-device AI tasks.
- 12-core CPU: 2 super cores, 4 performance cores, 6 efficiency cores
- 12-core GPU with hardware ray tracing and Dynamic Caching
- Dual 16-core Neural Engine for up to 2x AI compute over previous chips
- Up to 32GB unified memory and 170GB/s bandwidth, 10 percent above the M5
- 50 percent higher geometry rates for complex graphics and gaming
Apple also pitched the M6 GPU at gaming, pairing a redesigned shader core with hardware-accelerated ray tracing and second-generation Dynamic Caching to push higher frame rates and richer scene detail.
The new Mac mini keeps its supersmall footprint while gaining the M6 at entry level and an M5 Pro option at the top. Apple says the design balances the everyday speed users expect with machine that a growing number will run for local AI work rather than cloud calls.
A Bigger Play for On-Device AI
The M6 Mac mini targets a wider curve than past desktop refreshes, from everyday users and students to developers who want local LLM inference without the cloud. Faster prompt processing and a wider memory pipe make agentic tasks feel closer to native than ever before.
For heavy pro workloads, the M5 Ultra takes the spotlight in the Mac Studio. Built by fusing two dual-die M5 Max chips into Apple's first quad-die design, it scales to an 80-core GPU, 36-core CPU, 512GB of unified memory, and 1.2TB/s of bandwidth, enough to run hundred-billion-parameter models entirely on device.
The M6 is the first Apple chip where the CPU, GPU, and Neural Engine were all designed around the same job: running AI locally at speed. With the 2nm process now in volume and the Mac mini shipping at $899, Apple has put its on-device AI bet on store shelves.
For shoppers, the decision is simpler than the roadmap drama suggested. The M6 delivers a measurable step forward in both raw speed and local AI capacity today, and the M5 Ultra exists for anyone whose workflows need desktop-class memory beyond what a laptop can carry.
Comments