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?

Mars' North Polar Ice Is Much Cleaner Than Scientists Thought

https://scitechdaily.com/mars-north-polar-ice-is-much-cleaner-than-scientists-thought/
1•snarky-comments•1m ago•0 comments

Extra Big Ass Intelligence

https://www.extrabigassintelligence.com/
1•34679•2m ago•0 comments

Cloudflare Ohttp Gateway

https://blog.cloudflare.com/announcing-cloudflare-ohttp-gateway/
1•est•6m ago•0 comments

Show HN: Claude Scrolls TikToks for Me

https://tryrevline.com/
1•ZuraMakaradzeHe•6m ago•0 comments

Delta Is the Only Big Four Airline That Won't Use Starlink–Elon Musk Is Furious

https://www.wsj.com/business/airlines/delta-is-the-only-big-four-airline-that-wont-use-starlinkan...
2•doener•10m ago•0 comments

Lean Game Server: A repo of learning games for Lean

https://adam.math.hhu.de/
2•crescit_eundo•12m ago•1 comments

Hanami, Why?: Bits and Bobs

https://aaronmallen.me/writing/hanami-why-bits-bobs
1•thunderbong•19m ago•0 comments

Ask HN: Will source code become expensive if developers stop using GitHub?

3•debamitro•21m ago•1 comments

The Void: From Zero-Byte Responses to Continuation Control

https://zenodo.org/records/23070524
2•rayanpal_•24m ago•0 comments

Show HN: Fakeflac-go – A tool for quickly finding "fake" .flac files

https://github.com/drichline/fakeflac-go
1•DakotaR•25m ago•0 comments

Show HN: Hall Monitor. See all your agents across machines on your Mac's Notch

https://github.com/hiteshbandhu/hallmonitor
2•HiteshBandhu•26m ago•1 comments

SoftServe: A Scalable Quasi-Newton Method for Deep Learning

https://arxiv.org/abs/2610.02182
1•E-Reverance•34m ago•0 comments

Is Claude Conscious?

https://www.nytimes.com/2026/09/29/us/anthropic-claude-morals-ai.html
3•Anon84•37m ago•2 comments

Influencers are the private equity of culture

https://saturation.social/@jkottke/117370076406055289
2•colinprince•39m ago•0 comments

How to stay smart in the age of AI: the science of critical thinking

https://www.nature.com/articles/d41586-026-02930-6
1•zzzeek•42m ago•0 comments

How fast can a computer-use agent finish the job?

https://cuaspeedrun.com/
2•handfuloflight•42m ago•0 comments

Academic Doomerism

http://muratbuffalo.blogspot.com/2026/09/academic-doomerism.html
1•greghn•44m ago•0 comments

Humanoid robots won't surprise us when they arrive

https://philipotoole.com/humanoid-robots-wont-surprise-us-when-they-appear/
2•otoolep•48m ago•0 comments

Tavus claims 48% pass rate if video Turing test

https://twitter.com/tavus/status/2105704169009246248
1•CoryOndrejka•51m ago•0 comments

New York City should carefully measure a new tree

https://blog.willmeye.rs/new-york-city-should-carefully-measure-a-new-tree/
1•willmeyers•53m ago•0 comments

MTurk closed this week: how to keep Ground Truth and A2I workflows running

https://services.deepen.ai/guides/mturk-ground-truth-a2i-migration
1•mmusa•54m ago•0 comments

Prenatal exposure to the plasticizer DEHP increases autism and ADHD

https://www.sciencedirect.com/science/article/pii/S2666634026002941
2•OutOfHere•54m ago•2 comments

AmiGalaga: Galaga for the stock A500 at 50 FPS (OCS, 512K), open source

https://www.reddit.com/r/amiga/comments/1ww5jaf/amigalaga_galaga_for_the_stock_a500_at_50_fps_ocs/
1•doener•56m ago•0 comments

Go Mqtt Cluster Tracker

https://github.com/AUTOSOLN/swarmy-mqtt
1•redbeard281•59m ago•0 comments

Do Less, So You Can Do Better

https://www.indiehackers.com/post/do-less-so-you-can-do-it-better-3ea77e92b4
4•donchuru•59m ago•0 comments

V8.16 Global Optimal Parser – Lossless Data Compression

https://gitlab.com/4d-solutions-gruppe/project-v816-parser/-/blob/main/README.md
1•Amaniel•1h ago•0 comments

ThreeJS Port of OpenDLSS-NR

https://github.com/bhouston/three-dlss-nr/
1•bhouston•1h ago•0 comments

Agentic Robotics Solves an Industrial Production Problem

https://www.ambirobotics.com/blog/agentic-robotics/
1•gmays•1h ago•0 comments

Gloss – Annotate live code and provide feedback to an agent

https://github.com/ryanbrunner/gloss
2•ryanbrunner•1h ago•1 comments

ArXiv imposes rate limit on paper submissions to stem the AI slop tide

https://www.theregister.com/ai-and-ml/2026/10/02/arxiv-imposes-rate-limit-on-paper-submissions-to...
4•sbulaev•1h ago•2 comments