Expert Checksum with AVX2
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
public static unsafe uint ChecksumExpertAvx2(ReadOnlySpan<byte> arr) | |
{ | |
ref byte refSpan = ref MemoryMarshal.GetReference<byte>(arr); | |
var z = 0; | |
uint sum = 0; | |
var vectorSum = Avx2.Xor(Vector256<byte>.Zero, Vector256<byte>.Zero).AsUInt32(); | |
var mask = Vector256.Create((byte)3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12, (byte)3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12); | |
int limit = arr.Length - 128; | |
while (z <= limit) | |
{ | |
var v1 = Vector256.LoadUnsafe(ref Unsafe.Add(ref refSpan, z)); | |
var v2 = Vector256.LoadUnsafe(ref Unsafe.Add(ref refSpan, z + 32)); | |
var v3 = Vector256.LoadUnsafe(ref Unsafe.Add(ref refSpan, z + 64)); | |
var v4 = Vector256.LoadUnsafe(ref Unsafe.Add(ref refSpan, z + 96)); | |
var s1 = Avx2.Shuffle(v1, mask).AsUInt32(); | |
var s2 = Avx2.Shuffle(v2, mask).AsUInt32(); | |
var s3 = Avx2.Shuffle(v3, mask).AsUInt32(); | |
var s4 = Avx2.Shuffle(v4, mask).AsUInt32(); | |
var s5 = Avx2.Add(Avx2.Add(s1, s2), Avx2.Add(s3, s4)); | |
vectorSum = Avx2.Add(vectorSum, s5); | |
z += 128; | |
} | |
... | |
} |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment