Re:Zen4's AVX512 Teardown
by adrian_b on Monday September 26, 2022 @06:01PM
(#32989133)
I assume that as you say, whoever used this instruction was using it for multiplying 32-bit numbers.
On AMD Zen 4 and Intel Cannon Lake or newer (when AVX-512 is supported), the fastest method to multiply big numbers is to use the IFMA instructions, which reuse the floating-point multipliers to generate 104-bit products of 52-bit numbers.
On AMD Zen 4 and Intel Cannon Lake or newer (when AVX-512 is supported), the fastest method to multiply big numbers is to use the IFMA instructions, which reuse the floating-point multipliers to generate 104-bit products of 52-bit numbers.