Open amitdo opened 2 years ago
... for int dot product.
The equivalent for VNNI VPDPBUSD seems to be USDOT.
VPDPBUSD
USDOT
https://developer.arm.com/documentation/ddi0596/2021-12/SIMD-FP-Instructions/USDOT--by-element---Dot-Product-with-unsigned-and-signed-integers--vector--by-element--
There are some variants: SDOT, SUDOT, UDOT.
SDOT
SUDOT
UDOT
CC: @robinwatts
Arm's USMMLA instruction seems even better.
USMMLA
https://developer.arm.com/documentation/ddi0596/2021-12/SIMD-FP-Instructions/USMMLA--vector---Unsigned-and-signed-8-bit-integer-matrix-multiply-accumulate--vector--
I wonder if Intel has an equivalent instruction.
... for int dot product.