Open nigeltao opened 5 years ago
Actually, 14x should probably be 41x. See https://fastcompression.blogspot.com/2019/03/presenting-xxh3.html?showComment=1552696407071#c3490092340461170735
C/C++ x86 SIMD (SSE4.2 + PCLMUL) CRC-32 implementation: https://github.com/nigeltao/smhasher/commit/9f561bb6ceed7f884aff59a028fbfaff13825b2e
src/crc.cpp
is a very simple CRC-32 implementation. SIMD implementations can be more than 14x faster.See https://fastcompression.blogspot.com/2019/03/presenting-xxh3.html?showComment=1552693962180#c8263190774970267162