The heart of an AI processor: Is it the NPU and GPU?

Yes, for most situations.

January 26, 2026

Jon Peddie

When it comes to AI processing, there’s no single “heart”—it’s more like a family of specialized organs. GPUs, led by Nvidia’s powerhouse chips, dominate the heavy lifting of training massive AI models in data centers. Meanwhile, NPUs quietly rule the edge, powering everything from smartphone cameras to car sensors with incredible efficiency. But the plot thickens: Many companies build NPU-like chips but give them unique names—Apple calls theirs the Neural Engine, Google has TPUs, Groq branded the LPU, and AMD went with XDNA AI Engine. It’s a diverse ecosystem where each architecture shines in its own role. The AIP food

...

Enjoy full access with a TechWatch subscription!

TechWatch is the front line of JPR information gathering service, comprising current stories of interest to the graphics industry spanning the core areas of graphics hardware and software, workstations, gaming, and design.

A subscription to TechWatch includes 4 hours of consulting time to be used over the course of the subscription.

Already a subscriber? Login below

The heart of an AI processor: Is it the NPU and GPU?

Enjoy full access with a TechWatch subscription!

This content is restricted

Login