frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Show HN: DualMix128 – A fast, simple PRNG passing PractRand (32TB) and BigCrush

https://github.com/the-othernet/DualMix128
5•the_othernet•1y ago
Hi HN,

I'd like to share DualMix128, a pseudo-random number generator I developed. It's written in C and achieves very high speed while maintaining strong statistical properties for non-cryptographic use.

GitHub (MIT License): https://github.com/the-othernet/DualMix128

Key points:

* *Fast:* Benchmarked at ~0.36 ns per 64-bit generation on GCC 11.4 (-O3 -march=native). This was over 2x faster than `xoroshiro128++` (~0.74 ns) and competitive with `wyrand` (~0.36 ns) in the same tests. (Benchmark code/results in repo).

* *Robust:* Passes the full TestU01 BigCrush suite and PractRand up to 32TB with no anomalies reported. (Test harnesses/logs in repo).

* *Simple Algorithm:* Uses a 128-bit state (two `uint64_t`) with simple mixing based on addition, XOR, rotation, and a final multiplication by the golden ratio constant.

* *C Implementation:* Minimal dependencies (`stdint.h` for core), straightforward C code.

This came out of exploring simple constructions for efficient PRNGs that still hold up under heavy statistical testing.

Curious to hear feedback on the design, performance, potential applications, or any weaknesses you might spot. Thanks!

Comments

thomaskoopman•1y ago
Very cool, fast and looks like it should vectorize too. Do you have a jump function for parallel seeding?

How did you come up with this, some number-theoretic basis or more experimental?

Why the Largest Megaproject Collapsed` [video]

https://www.youtube.com/watch?v=GbB3vFfiaVU
1•thelastgallon•41s ago•0 comments

Open source FPGA: the next frontier is silicon

https://www.reddit.com/r/FPGA/comments/1wbrrgi/open_source_fpga_the_next_frontier_is_silicon/
1•random__duck•6m ago•1 comments

FlavorDB: A resource to explore flavor molecules

https://cosylab.iiitd.edu.in/flavordb/
2•thunderbong•6m ago•0 comments

AI Made Delegation Easy: Structure Is Still the Skill

https://syntheticauth.ai/posts/ai-made-delegation-easy-structure-is-still-the-skill
1•zerolayers•12m ago•0 comments

Kepler Compute- New memory startup for AI

https://keplercompute.com
1•osnium123•12m ago•0 comments

Siri Recap: AI Memory with Apple Watch

https://www.apple.com/hk/en/apple-watch-series-12/
1•chepy•18m ago•0 comments

Carmichael numbers are quickly findable (Lean-verified) [pdf]

https://jdb19937.github.io/carmichael/carmichael.pdf
1•jdb1729•18m ago•1 comments

Show HN: They chose your competitor. Find out why

https://drop.space
3•zhangchen0411•25m ago•0 comments

Experimental Evaluation Methodology for the Era of No Steady Performance [pdf]

https://www.d3s.mff.cuni.cz/files/publications/antoch_experimental_2026.pdf
3•luu•26m ago•0 comments

Show HN: NYC rideshare market analysis (2019 – 2026)

https://github.com/omarirfa/rideshare_market
3•fl4res•26m ago•0 comments

Show HN: Find your next client on X

https://tiny.opencreators.ai
4•zhangchen0411•26m ago•0 comments

Hermeus Unveils Ramjet-X, a Low-Cost Test Vehicle for High-Mach Flight

https://www.hermeus.com/article/hermeus-unveils-ramjet-x
3•r2sk5t•28m ago•0 comments

IonQ secp256k1 Estimate: 20k Qubits, 26 Days per Try

https://postquantum.com/post-quantum/ionq-secp256k1-resource-estimate/
3•giuliomagnifico•31m ago•0 comments

Busabase for DeepSeek Harness: An Agent database that runs apps and skills

https://github.com/busabase/busabase-dsh-plugin
2•poppylorry•33m ago•0 comments

Is OpenAI the Achilles' Heel of the US Economy? [video]

https://www.youtube.com/watch?v=RhLLwfwlNAU
2•ferryth•33m ago•0 comments

Can LLM Agents Infer World Models? Evidence from Agentic Automata Learning

https://arxiv.org/abs/2606.16576
2•wslh•36m ago•0 comments

Training a 3.8B LLM to 0.384 CORE for $998 – Hugo Vergnes

https://hugovergnes.github.io/little-lm-3-8b/
2•Anon84•41m ago•0 comments

Prompt Trees: Training-Time Prefix Caching

https://www.scaledcognition.com/blog/prompt-trees
2•ankitg12•41m ago•0 comments

Designing a new conceptual package manager– tell me what's wrong

https://github.com/Rivis666/Leash
2•Rivis•44m ago•0 comments

A Stupid Idea for AI Alignment We Came with by Looking at Specification Gaming

https://slimemoldtimemold.com/2026/08/05/a-stupid-idea-for-ai-alignment-we-came-up-with-by-lookin...
2•cpeterso•45m ago•0 comments

Semi-Implicit Pairwise Descent for Nonlocal Continuum Mechanics

https://arxiv.org/abs/2609.09834
2•E-Reverance•46m ago•0 comments

Show HN: What if the speed of light was 5 km/h?

https://rivendell.dmitrybrant.com/relativity/
8•dmitrybrant•47m ago•5 comments

How we manage and engage with our horses shapes their personality

https://www.utu.fi/en/news/press-release/how-we-manage-and-engage-with-our-horses-shapes-their-pe...
4•thunderbong•47m ago•0 comments

RDC – Remote Desktop Control for AI Agents

https://github.com/bscott/rdc
3•blstech•54m ago•0 comments

How HN: Dashly – Cookieless analytics with revenue attribution and bot filtering

https://www.dashly.lol
2•Quise2003•55m ago•0 comments

Playing mario 64 using a fly's brain

https://twitter.com/barrelshifter/status/2097004115826200898
3•rguiscard•1h ago•2 comments

What Builders Made with GPT-6 Astra (So Far)

https://clipnotebook.com/p/103b93d0-65fe-4b3f-a4c3-a1d0275d29cc
3•diddddy•1h ago•0 comments

The threat of vibe-coded superviruses

https://www.noahpinion.blog/p/heres-how-were-all-going-to-die
1•ryan_j_naughton•1h ago•0 comments

California enacts AI safety evaluation laws backed by Anthropic and OpenAI

https://politico.com/news/2026/09/09/newsom-signs-ai-safety-bills-backed-by-anthropic-openai-0106...
2•utiiiD•1h ago•1 comments

My Rebuttal to the Anthropic Extinction Event Hype Train

3•hmichaelson24•1h ago•0 comments