Available on (x86 or x86-64) and target feature
fma
and x86-64 only.Expand description
Multiplies packed single-precision (32-bit) floating-point elements in a
and b
, and subtract packed elements in c
from the intermediate result.