Function _mm_sub_ps

#[target_feature(enable = "sse")]
pub const fn _mm_sub_ps(a: __m128, b: __m128) -> __m128

Subtracts packed single-precision (32-bit) floating-point elements in a and b.

Intel's documentation