Apple's latest high-end chips, the M5 Pro and M5 Max, introduce substantial changes in their architecture. Unlike previous generations, these chips are no longer monolithic, packing all cores into a single silicon die. Instead, they utilize an 'all-new Fusion Architecture,' which involves separating the CPU and GPU cores into distinct silicon dies that are then integrated into one chip. This innovative approach is reminiscent of the technology used to merge two Max chips into a single Ultra chip.
Both the M5 Pro and M5 Max are built on the same 18-core CPU die, but they differ in their GPU configurations. The M5 Pro utilizes a 20-core GPU die, while the M5 Max is equipped with a 40-core GPU die. As the memory controller is part of the GPU die, the Max offers greater memory bandwidth and supports higher memory capacities than the Pro model.
Another key development is the absence of 'efficiency' CPU cores in these chips. The M5 family's high-performance cores have now been rebranded as 'super' cores in macOS 26.3.1, while a new kind of 'performance' core replaces the original efficiency cores in the M5 Pro and M5 Max. The standard M5 retains its smaller, slower efficiency cores.
A comparison of specifications highlights these enhancements: the M5 Max includes up to 6 'super' cores, 12 'performance' cores, 40 GPU cores, and offers memory bandwidth up to 614 GB/s. Meanwhile, the M5 Pro features up to 6 'super' cores, 12 'performance' cores, and 20 GPU cores, with memory bandwidth of 307 GB/s.
For users, the M5 Pro and M5 Max represent expected iterative improvements typical of new Apple Silicon releases. However, those interested in technical specifics will find the M5 Max’s performance and its rebranded 'super' cores worth a deeper investigation, especially to clarify the distinction between what's rebranded and what's genuinely new—such as the new 'performance' cores, which are distinctly different from earlier efficiency cores.