Closed buybackoff closed 6 years ago
I used your code in my CRC32 implementation and did some benchmarks: https://github.com/Spreads/Spreads/commit/1a6b6c252989bf5071be23366a78429aa6751955
CRC32 Managed in every benchmark is code from this repo rewritten to avoid virtual method calls and works with pointers. The numbers are in MB/sec.
CRC32 Managed
The main point was to use hardware intrinsics and your code is the fastest managed fallback I have found.
Thanks!
I used your code in my CRC32 implementation and did some benchmarks: https://github.com/Spreads/Spreads/commit/1a6b6c252989bf5071be23366a78429aa6751955
CRC32 Managed
in every benchmark is code from this repo rewritten to avoid virtual method calls and works with pointers. The numbers are in MB/sec.The main point was to use hardware intrinsics and your code is the fastest managed fallback I have found.
Thanks!