Apple's new chips double down on local AI processing for Macs
Apple introduced the M5 Ultra and M6 chips for its new Mac Mini and Mac Studio. The M5 Ultra combines two M5 Max dies into a quad-die design with up to 36 CPU cores and 80 GPU cores, while the M6 uses a 2nm process with a 12-core CPU and dual 16-core Neural Engine. Both chips emphasize on-device AI performance, with the M6 offering nearly 30% faster peak GPU compute for AI than the M5.
EXPANDED:
The M5 Ultra's quad-die architecture represents a notable engineering shift, merging two M5 Max dies to reach 36 CPU cores and 80 GPU cores. Its 1.2 TB/s unified memory bandwidth marks a 50 percent increase over the M3 Ultra released roughly eighteen months earlier. This configuration targets professionals handling complex rendering, visual effects, and frontier AI model workloads.
The M6 chip, built on a 2nm process, adds two CPU cores over its predecessor for a 12-core layout, paired with a dual 16-core Neural Engine. Apple reports nearly 30 percent faster peak GPU compute for AI tasks, accelerating on-device prompt processing for local language models. The company's developer frameworks automatically distribute workloads