According to a Bloomberg "Power On" newsletter by Mark Gurman, Apple plans a six-month gap between the M6 processor debut and the AI-focused M7, with the base M7 arriving in the first half of 2027. The M6 Pro, M6 Max, and M6 Ultra will not exist; instead, Apple is focusing on AI-first chips for the M7 generation. The M6 will improve memory bandwidth from 153 gigabytes per second in the M5 to 200 gigabytes per second, using a new memory architecture, a boosted Neural Engine, and 12 GPU cores. The M7 generation will have a much bigger focus on AI processing, with base chip memory bandwidth increasing to around 240 gigabytes per second. The M7 Pro and M7 Max are expected at the end of 2027, followed by the M7 Ultra in 2028. Gurman suggests the M7 Ultra, supporting as much as 1.5 terabytes of memory, could be the basis for Apple's AI server strategy, with a new server chip built around it potentially appearing by 2029.