Using AVX to improve performance of float subtract, divide, truncate to int32

Question

Trying to use AVX to improve performance of the following

__declspec(dllexport) void __cdecl calculate_quantized_vertical_values(long length, float min, float