SHISHUA – The Fastest PRNG In The World

The announcement and explanation blog post is here.

Note: please do not use this for cryptographic purposes. If you need security, a recommended option is ChaCha20.

Implementations:

• C
• Rust
• Node.js (on npm)
• Python (on PyPI)

Comparison

<table> <tr><th>Name <th><a href='./test/benchmark-perf-intel'>Intel Performance</a> <th><a href='./test/benchmark-perf-arm'>ARM Performance</a> <th>Quality <th>Seed correlation <tr><td>SHISHUA <td>52.93 GB/s <td> 8.78 GB/s <td> >32 TiB <td> >32 TiB <tr><td>xoshiro256+x8 <td>40.18 GB/s <td> 3.69 GB/s <td> 1 KiB <td> 0 KiB <tr><td>RomuTrio <td> 8.62 GB/s <td> 4.76 GB/s <td> >32 TiB <td> 1 KiB <tr><td>xoshiro256+ <td> 7.88 GB/s <td> 4.31 GB/s <td> 512 MiB <td> 1 KiB <tr><td>wyrand <td> 7.00 GB/s <td> 2.53 GB/s <td> >32 TiB <td> 32 KiB <tr><td>Lehmer128 <td> 6.14 GB/s <td> 1.83 GB/s <td> >32 TiB <td> 1 KiB <tr><td>ChaCha8 <td> 6.28 GB/s <td> 1.72 GB/s <td> >32 TiB? <td> >32 TiB? <tr><td>RC4 <td> 0.35 GB/s <td> 0.15 GB/s <td> 1 TiB <td> 1 KiB </table>

1. Performance: in number of CPU cycles spent per byte generated, on N2 GCP instances. On N2D (AMD), the order is the same.
2. Quality: level at which it fails PractRand. We show a > if it did not fail. We put a question mark if we have not proved it.
3. Seed correlation: PractRand on interleaving of bytes from eight streams with seeds 1, 2, 4, 8, 16, 32, 64, 128. We use PractRand with folding 2 and expanded tests.

On the subject of seed correlations, the ./bin/sample-seed-fingerprints.sh program highlights additional anomalies:

• RC4 has weak seeds with heavy artefacts visible in the fingerprints.
• RomuTrio's fingerprints start with the same first character for all seeds.

Commands

• make: build ./shishua.
• make test: run performance tests, PractRand, and BigCrush on SHISHUA.
• make test/benchmark-seed: run seed correlation test.
• make test/benchmark-perf: run performance comparison locally.
• make benchmark-intel: run performance comparison on a GCP Intel chip.
• make benchmark-amd: run performance comparison on a GCP AMD chip.
• make benchmark-arm: run performance comparison on an AWS Graviton ARM chip.

---

The SHISHUA and SHISHUA-half are under the CC0 license.