Apple unveils M6 and M5 Ultra chips with major AI performance gains

Apple’s M6 chip, built on a 2nm manufacturing process, represents a significant shift in the company’s silicon roadmap, with on-device AI workloads now sitting at the center of its strategy. On the graphics side, the M6 is expected to feature a redesigned GPU with up to 12 cores and memory bandwidth of approximately 200 GB/s. For context, the M5 offered around 153 GB/s. That’s roughly a 30% jump in memory throughput. The 2nm process node is manufactured by TSMC. Shrinking transistors to this scale allows more of them to fit on the same die, which translates to better performance per watt. A strategic pivot toward AI-first silicon Reports indicate that Apple will not be releasing M6 Pro, M6 Max, or M6 Ultra variants. Instead, those higher-end slots in the product lineup will be filled by the M7 chip family, which is expected to arrive in 2027 with an explicitly AI-optimized architecture. The base M6 chip is expected to power refreshed entry-level products, including a 14-inch MacBook Pro and a new 24-inch iMac, with availability projected for late 2026. Apple has not provided official confirmations or unveiled detailed specifications as of late August 2026, leaving the industry spec...