Apple Unveils M5 Ultra and 2 nm M6 Chips for Local AI Macs

Summary

Apple announced the M5 Ultra and M6 on Tuesday, August 25, 2026, for new Mac Studio and Mac Mini models. Branded its most powerful chip yet, M5 Ultra combines two dual-die M5 Max chips into Apple’s first quad-die processor, with up to a 36-core CPU, 80-core GPU and 1.2 TB/s of unified memory bandwidth, 50% more than the roughly 18-month-old M3 Ultra. It targets professionals handling complex 3D rendering, visual effects and frontier AI models, while also serving hobbyist AI developers.

Built on a 2 nm process, M6 introduces a new CPU complex, additional CPU and GPU cores, a Dual 16-core Neural Engine and greater memory bandwidth, Apple silicon engineering vice president Sri Santhanam said. Its 12-core CPU has two more cores than M5, while peak GPU compute for AI rises nearly 30%, accelerating prompt processing for on-device LLMs. Apple’s tools optimize workloads across the CPU, GPU and Neural Engine, letting developers locally run and fine-tune Apple Foundation Models or proprietary models and access App Intents and Apple Intelligence. Local processing extends Apple’s privacy approach behind Face ID and Touch ID, although its long-awaited Siri upgrade uses Google Gemini as Apple trails rivals in proprietary AI. Mac Mini preorders start at $899, with shipments after September 22.