AMD & Intel’s ACE Partnership Significantly Boosts AI Performance as the “Standard Matrix-Acceleration Architecture” for x86 Chips

AMD and Intel published an ACE whitepaper outlining AI Compute Extensions for x86 chips.

ACE is a new matrix-multiply instruction-set extension designed to work with AVX10.

The goal is to improve AI performance, scalability, and energy efficiency across CPUs.

A shared x86 acceleration layer could reduce software fragmentation for developers and vendors.