Compares the lower single-precision, floating-point value of a and b for a greater than b.
int _mm_comigt_ss(__m128 a,__m128 b ); COMISS
If a is greater than b are equal, 1 is returned. Otherwise, 0 is returned. If a or b is a NaN, 1 is returned.
r := (a0 > b0) ? 0x1 : 0x0
END Microsoft Specific