Longtime rivals Intel and AMD are joining forces on a new set of CPU instructions designed for artificial intelligence work. According to Tom's Hardware, the two chipmakers have introduced extensions called ACE that bring an AI-oriented instruction set to the x86 architecture — the technology that powers most laptops, desktops, and servers.
The headline feature is a new design for matrix multiplication, the repetitive number-crunching that sits at the heart of modern AI models. Tom's Hardware reports that ACE makes this kind of math more power-efficient and more density-efficient, meaning chips can do the same work while drawing less energy and packing more capability into the same space.
That focus matters because matrix multiplication is one of the most demanding operations in machine learning, and doing it more efficiently is a central challenge for the whole industry. By building the capability directly into the x86 instruction set, Intel and AMD are aiming to make standard processors better suited to AI tasks rather than leaning solely on separate accelerators.
It is also notable that Intel and AMD, who compete head-to-head in the processor market, are aligning on a shared standard here. A common instruction set means software developers can write code once and have it run across chips from both companies, which tends to speed adoption.
Why it matters: if ACE delivers the efficiency gains Tom's Hardware describes, it could make the everyday processors inside ordinary computers meaningfully better at running AI — without the higher power costs that usually come with it.