blake2b-benchmark

Benchmark of various Blake2b implementations.

To run the benches:

npm install
npm run bench

The benchmark tests how long it takes to hash a series of 128b, 4kb and 64kb buffers using sodium-native, blake2b, and blake2b-wasm

Here are the results from running the bench on my 2015, MacBook 12".

Using sodium-native (Node.js native module)

NANOBENCH version 2
> node index.js sodium-native

# hash 100000x 128b
ok ~115 ms (0 s + 115110995 ns)

# hash 1000x 4kb
ok ~6.09 ms (0 s + 6093677 ns)

# hash 1000x 64kb
ok ~101 ms (0 s + 100862480 ns)

all benchmarks completed
ok ~222 ms (0 s + 222067152 ns)

Using blake2b-wasm (WebAssembly)

NANOBENCH version 2
> node index.js blake2b-wasm

# hash 100000x 128b
ok ~222 ms (0 s + 221547739 ns)

# hash 1000x 4kb
ok ~22 ms (0 s + 21884462 ns)

# hash 1000x 64kb
ok ~306 ms (0 s + 305651336 ns)

all benchmarks completed
ok ~549 ms (0 s + 549083537 ns)

Using blake2b (Pure Javascript)

NANOBENCH version 2
> node index.js blake2b

# hash 100000x 128b
ok ~1.01 s (1 s + 12673122 ns)

# hash 1000x 4kb
ok ~327 ms (0 s + 326822314 ns)

# hash 1000x 64kb
ok ~5.25 s (5 s + 254090932 ns)

all benchmarks completed
ok ~6.59 s (6 s + 593586368 ns)

License

MIT