Fastest hashing algorithm. It is proposed in four flavors (XXH32, XXH64, XXH3_64bits and XXH3_128bits). If you're only comparing two files once, faster than any hash algorithm will be simple reading the files and comparing them. Suitable for file and large data hashing (checksums). Clang-18+ compiles it to ~185 instructions, both on x86-64 and aarch64. The library includes the following algorithms : XXH32 : generates 32-bit hashes, using 32-bit arithmetic XXH64 : generates 64-bit hashes, using 64-bit arithmetic XXH3 (since v0. But the OP just wants to store values in a hashtable, and I don't think a cryptographic hash function is really appropriate for that. Surpasses 70GB/s on Apple's M4 cpus. This benchmark was done on 100 MB of random data, with each hashing algorithm hashing the same random data blob, averaged from 5 runs, on an isolated Ubuntu 20. Contribute to google/highwayhash development by creating an account on GitHub. This performance is possible mostly thanks to heavy usage of SIMD intrinsics, high ILP construction, a small bytecode (easily inlined and cached) and some (outrageously unsafe) tricks. 0eqgk ydkac 7ud27v dfbvxe stto jx7 io yzn xgnfkyek uch