pub fn f16_mul(a: u16, b: u16) -> u16
One correctly-rounded IEEE-754 binary16 multiply (see f16_add; the product of two f16s is exact in f32).
f16_add