Neural Engine
Apple's dedicated on-chip block for AI work, separate from the main processor and graphics cores.
The Neural Engine is a section of Apple’s chips built for one job: the repetitive multiply-and-add arithmetic that machine learning models are made of. It handles that far more efficiently than the general-purpose CPU and with less power draw than the graphics cores, which is why on-device features like dictation, photo search and text prediction can run all day without flattening the battery. Other companies call their version an NPU.
Apple rates it in cores, and the M6 announced in August 2026 introduced a dual 16-core arrangement for roughly twice the peak AI throughput of the previous generation. Apple has also started adding smaller “Neural Accelerators” inside individual GPU cores, so the AI work is spread across the chip rather than confined to one block.