Apple's base MacBook Pro is getting an M6 chip later this year — the company's first built on a 2nm process.
According to Bloomberg, Apple plans to launch the M6 in late 2026, starting with the 14-inch MacBook Pro. The chip will use TSMC's N2 process node, which packs more transistors into the same die area compared to the current 3nm design. Memory bandwidth jumps from 153GB/s on the M5 to roughly 200GB/s, and the GPU grows from 10 cores to 12. Apple is also moving to a new packaging technology — WMCM, or Wafer-Level Multi-Chip Module — which places the CPU, GPU, DRAM, and Neural Engine in closer physical proximity to cut latency between them.
The bandwidth increase matters more than the headline node shrink. AI inference and high-resolution graphics are both memory-starved workloads, and a 30 percent bandwidth gain compounds across every task that hits the GPU or Neural Engine. Apple's bet on tighter chip integration, rather than raw core counts, is the quiet story here — and it's a direct answer to the memory bottlenecks that have constrained on-device AI on existing silicon.
There is a notable gap in the lineup, though. Apple is not building M6 Pro or M6 Max chips; those arrive with the M7 in 2027. Anyone waiting for a high-end MacBook Pro upgrade will sit out this cycle. The Mac mini, iMac, and MacBook Air could also see M6 updates, but the report only names the MacBook Pro specifically — so treat the rest as possibilities, not plans.