Benchmark of Rust PRNG crates

Simple performance benchmark of pseudorandom generation Rust crates for non-cryptographic purposes such as games.

The goal is not to compare algorithms, but implementations that can be used immediately by a Rust developer.

Inclusion criteria:

• Seedable. Needed for reproducibility.
• Simple to use, either thanks to an obvious API or examples in documentation.
• Works on common targets.
• Usable for generating bytes or groups of bytes.
• Builds on my system.

Warning: If your need has anything to do with security, i.e. generating values seeded by or interacting with potentially untrusted data or that should be secret or not leak information about the seed or correlation between multiple outputs, then you should instead use a CSPRNG.

Results

RUSTFLAGS='-C target-cpu=native' cargo run --release

rustc 1.95.0-nightly (57d2fb136 2026-02-01)

Results on AMD Ryzen 7 5700X: (with AVX2)

Time (s)        PRNG
0.06720725      simd_rand::specific::avx2::xoshiro256plusplus::Xoshiro256PlusPlusX4
0.06825219      simd_rand::portable::xoshiro256plusx4::Xoshiro256PlusX4
0.06890218      simd_rand::specific::avx2::xoshiro256plus::Xoshiro256PlusX4
0.069135666     simd_rand::portable::xoshiro256plusplusx4::Xoshiro256PlusPlusX4
0.07073431      simd_rand::portable::xoshiro256plusx8::Xoshiro256PlusX8
0.07283282      simd_rand::portable::xoshiro256plusplusx8::Xoshiro256PlusPlusX8
0.07305191      frand::Rand
0.076635174     fast_rands::RomuDuoJrRand
0.07750931      rand_xoshiro::xoshiro256plus::Xoshiro256Plus
0.0798837       wyhash::v1::traits::WyRng
0.08170648      fast_rands::RomuTrioRand
0.082019165     romu::Rng
0.08613385      simd_rand::specific::avx2::shishua::Shishua<32768>
0.08767373      tinyrand::wyrand::Wyrand
0.08789389      wyhash::final3::traits::WyRng
0.087896034     fastrand::Rng
0.08798213      wyrand::final_v4_2::wyrand::WyRand
0.089724645     rand_xoshiro::xoshiro256starstar::Xoshiro256StarStar
0.090424806     rand_xoshiro::xoroshiro128plus::Xoroshiro128Plus
0.09778529      tiny_prng::pcg::PcgXslRr6432Mcg
0.098746814     rand_xoshiro::xoroshiro128plusplus::Xoroshiro128PlusPlus
0.10192768      rand_xoshiro::xoroshiro128starstar::Xoroshiro128StarStar
0.10222766      fast_rands::Xoshiro256PlusPlusRand
0.1025012       rand_xoshiro::xoshiro256plusplus::Xoshiro256PlusPlus
0.10895407      rand::rngs::small::SmallRng
0.109856725     rand_xoshiro::xoshiro512plus::Xoshiro512Plus
0.1161916       tinyrand::xorshift::Xorshift
0.11952907      shishua::core::ShiShuAState
0.1273132       tiny_prng::pcg::PcgXslRrMcg
0.1313371       tiny_prng::xorshift::Xorshift1024star
0.13799635      tiny_prng::xorshift::Xorshift128
0.1382492       rand_xoshiro::xoshiro512plusplus::Xoshiro512PlusPlus
0.13890779      rand_xoshiro::xoshiro512starstar::Xoshiro512StarStar
0.14007749      rand_xoshiro::splitmix64::SplitMix64
0.14576794      rand_xoshiro::xoshiro128plus::Xoshiro128Plus
0.15051891      fast_rands::Sfc64Rand
0.1522026       tiny_prng::pcg::PcgXslRr
0.15726987      tinyrand::splitmix::SplitMix
0.15869391      prng_mt::mt19937::MT19937_64
0.15959476      rand_xoshiro::xoshiro128starstar::Xoshiro128StarStar
0.1698655       bsd4random::BSD
0.17368329      tiny_prng::xorshift::Xorshift64
0.17383604      rand_xoshiro::xoroshiro64star::Xoroshiro64Star
0.17391294      fast_rands::XorShift64Rand
0.1766414       fast_rands::Lehmer64Rand
0.17737824      tiny_prng::xorshift::Xorshift64star
0.18385774      mcg59::MCG59
0.18856412      rand_xoshiro::xoshiro128plusplus::Xoshiro128PlusPlus
0.20034873      tiny_prng::mt64::Mt19937
0.21738203      rand_xoshiro::xoroshiro64starstar::Xoroshiro64StarStar
0.21857128      aes_prng::AesRng
0.23149101      tiny_prng::pcg::PcgXshRs6432
0.23159662      tiny_prng::pcg::PcgXshRr6432
0.28765705      rand::rngs::std::StdRng
0.29028276      prng_mt::mt19937::MT19937
0.3162913       quad_rand::RandGenerator
0.34691545      tiny_prng::xorshift::Xorshift32
0.34718052      nanorand::rand::wyrand::WyRand
0.3953146       tiny_prng::mt::Mt19937
0.40042716      nanorand::rand::pcg64::Pcg64
0.9632352       nanorand::rand::chacha::ChaCha<8>
1.772863        shishua::rand::ShiShuARng
3.5408454       acorn_prng::Acorn
3.706891        simplerand::Random

Results on Intel Core i5-7300HQ: (with AVX2)

Time (s)	PRNG
0.08826336	frand::Rand
0.0886573	simd_rand::specific::avx2::xoshiro256plusplus::Xoshiro256PlusPlusX4
0.08876129	simd_rand::portable::xoshiro256plusplusx4::Xoshiro256PlusPlusX4
0.089277595	simd_rand::portable::xoshiro256plusx4::Xoshiro256PlusX4
0.09128572	simd_rand::specific::avx2::xoshiro256plus::Xoshiro256PlusX4
0.09481453	simd_rand::portable::xoshiro256plusx8::Xoshiro256PlusX8
0.10730715	fast_rands::RomuDuoJrRand
0.10897806	fast_rands::RomuTrioRand
0.10994133	simd_rand::specific::avx2::shishua::Shishua<32768>
0.11030309	romu::Rng
0.122153305	simd_rand::portable::xoshiro256plusplusx8::Xoshiro256PlusPlusX8
0.12979501	rand_xoshiro::xoshiro256plus::Xoshiro256Plus
0.13618393	rand_xoshiro::xoshiro256starstar::Xoshiro256StarStar
0.14781812	rand_xoshiro::xoroshiro128plus::Xoroshiro128Plus
0.15827414	fast_rands::Xoshiro256PlusPlusRand
0.15854521	rand_xoshiro::xoshiro256plusplus::Xoshiro256PlusPlus
0.1587491	rand::rngs::small::SmallRng
0.16304855	shishua::core::ShiShuAState
0.1643057	rand_xoshiro::xoroshiro128plusplus::Xoroshiro128PlusPlus
0.16563393	tinyrand::wyrand::Wyrand
0.1673208	fastrand::Rng
0.16902256	wyhash::v1::traits::WyRng
0.16907188	wyrand::final_v4_2::wyrand::WyRand
0.16940956	wyhash::final3::traits::WyRng
0.17095926	rand_xoshiro::xoroshiro128starstar::Xoroshiro128StarStar
0.17670166	fast_rands::Sfc64Rand
0.17687643	tinyrand::xorshift::Xorshift
0.18041892	rand_xoshiro::splitmix64::SplitMix64
0.18187234	tiny_prng::xorshift::Xorshift1024star
0.18540086	tiny_prng::pcg::PcgXslRrMcg
0.18792455	tiny_prng::xorshift::Xorshift128
0.19410424	rand_xoshiro::xoshiro512plus::Xoshiro512Plus
0.20391174	fast_rands::Lehmer64Rand
0.2041288	mcg59::MCG59
0.21445897	rand_xoshiro::xoshiro512plusplus::Xoshiro512PlusPlus
0.21997462	tinyrand::splitmix::SplitMix
0.22679603	rand_xoshiro::xoshiro512starstar::Xoshiro512StarStar
0.23136179	bsd4random::BSD
0.23638329	fast_rands::XorShift64Rand
0.23744445	tiny_prng::pcg::PcgXslRr
0.23768099	tiny_prng::xorshift::Xorshift64star
0.23874637	tiny_prng::xorshift::Xorshift64
0.2519162	tiny_prng::pcg::PcgXslRr6432Mcg
0.2522125	rand_xoshiro::xoshiro128plus::Xoshiro128Plus
0.26618654	rand_xoshiro::xoshiro128starstar::Xoshiro128StarStar
0.31676543	rand_xoshiro::xoshiro128plusplus::Xoshiro128PlusPlus
0.32150853	tiny_prng::pcg::PcgXshRs6432
0.3222107	prng_mt::mt19937::MT19937_64
0.3227826	aes_prng::AesRng
0.32352522	rand_xoshiro::xoroshiro64star::Xoroshiro64Star
0.3316228	tiny_prng::pcg::PcgXshRr6432
0.3506137	rand_xoshiro::xoroshiro64starstar::Xoroshiro64StarStar
0.42379355	nanorand::rand::wyrand::WyRand
0.42504936	quad_rand::RandGenerator
0.46230927	rand::rngs::std::StdRng
0.4671986	tiny_prng::xorshift::Xorshift32
0.5528513	tiny_prng::mt64::Mt19937
0.6312231	prng_mt::mt19937::MT19937
0.6692442	nanorand::rand::pcg64::Pcg64
1.2569348	nanorand::rand::chacha::ChaCha<8>
1.4407393	tiny_prng::mt::Mt19937
2.0606635	shishua::rand::ShiShuARng
3.7855747	acorn_prng::Acorn
15.568359	simplerand::Random

FRand is among the fastest, however it has no particular mathematical justification beyond the fact its parameters were found empirically to minimize bias.

simd_rand provides the fastest implementations of common PRNGs, both portable and SIMD-specific.

fast_rands implements RomuDuoJr, the fastest after simd_rand. It is part of the Romu family which has been studied. It provides good quality up to at least 2^48 bytes. However this crate has too many dependencies that should not be necessary if you only want a PRNG, so copying the wanted functions can be an option.

Surprisingly, the AES PRNG is quite fast, probably due to the AES instruction set.

Also to my surprise, Shishua was slower than RomuTrio, contrarily to the author's claim. Maybe that's the implementation's fault or my lack of AVX512, I should investigate. The streaming overhead can be significant, as we see with ShiShuARng which is very slow compared to the bare ShiShuAState. simd_rand's implementation is quite faster.