Apple unveils M6 and M5 Ultra chips, turning Macs into desktop AI powerhouses
创业邦 · wechat · 2026-08-26
Apple launched its first 2nm chip, the M6, and the new quad-die M5 Ultra, alongside updated Mac mini and Mac Studio lines, significantly boosting local AI compute power.
M6 Chip & Mac mini:
- Process Breakthrough: The M6 is Apple's first 2nm chip, featuring a 12-core CPU, 12-core GPU, and 170GB/s bandwidth.
- AI Performance: AI performance is up to 4x faster than the M4 model. LM Studio prompt processing is up to 4.8x faster (and up to 13.5x faster vs. the first-gen M1).
- Positioning Shift: Apple markets the Mac mini as an 'Always-on Agentic Device', acting as a desktop Agent server.
- Price: Starting price rises to 6999 RMB.
M5 Ultra & Mac Studio:
- Architecture: First Quad-Die design using UltraFusion to connect four dies.
- Extreme Specs: Up to 36-core CPU, 80-core GPU, 512GB unified memory, and 1.2TB/s bandwidth. AI peak performance hits 4.3x the M3 Ultra.
- Distributed Inference: Supports clustering multiple Mac Studios for distributed AI inference, with 4-node performance reaching 3x a single machine.
- Supply Constraints: Due to global memory shortages, the 512GB configuration is delayed until late October, and purchase limits are in place.
Related event: Apple Launches M6 and M5 Ultra Chips with 512GB Memory Mac Studio(12 posts)→
More from Embodied
- KLTNet: Learning-Based Sparse Feature Tracking for Improved Monocular VIO — zhenjun_zhao · 2026-08-26
- KUKA Robot Demo: Keeping Tool Center Point Still — Olivier__OG · 2026-08-26
- Dreame pivots to embodied AI among 4 core businesses; XPeng's He says humanoid robots may beat cars on margin — 创业邦 · 2026-08-26
- LimX gathers global robotics voices at WRC 2026 to discuss life beyond demos — broodsugar · 2026-08-26
- Tesla upgrades in-car Grok to world's #1 speech-to-speech AI — XFreeze · 2026-08-26
- Deepak Pathak Demonstrates In-Context Learning for Robotics with One-Shot Video Prompts — deepakpathak · 2026-08-26