binary search but through SIMD

support u8, i8, u16, i16, u32, i32, u64, i64

waiting for Benchmarks

size	std	sse	avx2
8bit * 4
8bit * 16
8bit * 128
16bit * 4
16bit * 16
16bit * 128
16bit * 512
16bit * 2048
16bit * 8192
32bit * 4
32bit * 16
32bit * 128
32bit * 512
32bit * 2048
32bit * 8192
64bit * 4
64bit * 16
64bit * 128
64bit * 512
64bit * 2048
64bit * 8192