int_hash 0.2.0

Very fast, very simple hash algorithm designed for use in integer hash maps & sets.
Documentation

int_hash crates.io Documentation

Very fast, very simple hash algorithm designed for use in integer hash maps & sets.

This crate attempts to provide the fastest option for integer key hashmaps in the Rust language. So the algorithm may change if a better method is found for this use case.

use int_hash::IntHashMap;
let mut map: IntHashMap<u32, &str> = IntHashMap::default();
map.insert(22, "abc");

When hashing data larger than 64-bits the hasher will fallback to a secondary algorithm suitable for arbitrary data (defaults to FxHasher).

Benchmark Performance

For more info see the the full benchmark report.

Hash Algorithm Integer Sample Set int_hash is
Rust default aka SipHash Random numbers 1.23-9.61x faster
Rust default aka SipHash Natural numbers 3.82-21.6x faster
Rust default aka SipHash 32× table 1.59-4.94x faster
fnv Random numbers 0.99-1.58x faster
fnv Natural numbers 2.11-11.08x faster
fnv 32× table 0.57-1.09x faster
rustc-hash aka fx Random numbers 0.95-1.29x faster
rustc-hash aka fx Natural numbers 1.26-2.15x faster
rustc-hash aka fx 32× table 0.94-1.28x faster
seahash Random numbers 1.16-5.57x faster
seahash Natural numbers 3.65-19.35x faster
seahash 32× table 1.33-3.11x faster
twox_hash aka xx Random numbers 1.25-9.55x faster
twox_hash aka xx Natural numbers 4.05-23.95x faster
twox_hash aka xx 32× table 1.68-5.67xx faster

Note: For > 64-bit keys int_hash will perform inline with rustc_hash.

Limitations

The algorithm is non-cryptographic.

Why is it so fast

int_hash is dedicated at solving integer-sized hashing. Producing a unique u64 from an integer is not a very difficult problem, though getting a good spread of values to minimise hashmap collisions is a little harder.

The current implementation uses simple u64 XOR mixing to spread values. The sheer simplicity of this approach makes the hashing operation very fast and the primitive spreading is good enough to produce best-in-class hashmap performance.