frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

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•7mo 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•7mo 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?

The Vatican returns sacred artifacts held for a century to Indigenous owners

https://www.cnn.com/2025/12/08/europe/vatican-pope-indigenous-artifacts-canada-intl-hnk
1•sipofwater•44s ago•1 comments

Ask HN: Async updates: Input friction vs. Output friction?

1•sangkwun•48s ago•0 comments

Show HN: I Built an AI platform for trainers to manage workouts and diets

https://extremefitness.app/
1•shuvrokhan•1m ago•0 comments

New Jersey looks to create hallucinogenic mushroom research program

https://www.njspotlightnews.org/2025/12/nj-looks-to-create-hallucinogenic-mushroom-research-program/
1•magictruffle•2m ago•0 comments

Ask HN: Which domain name registrar do you use?

1•not-so-darkstar•4m ago•1 comments

Show HN: Web app that lets you send email time capsules

https://resurf.me
2•walrussama•6m ago•0 comments

Good Conversations Have Lots of Doorknobs (2022)

https://www.experimental-history.com/p/good-conversations-have-lots-of-doorknobs
1•bertwagner•7m ago•0 comments

Show HN: ZeroNotes – Client-side encrypted notes (AES-256-GCM and Argon2id)

https://app.zeronotes.is
2•Bjoern_Dev•9m ago•0 comments

Why U.S. senator in top Intel post wants more spying on Chinese companies

https://www.cnbc.com/2025/12/06/china-us-technology-spying-senate-concerns.html
1•ironyman•14m ago•0 comments

Survivors Clung to Wreckage for Some 45 Minutes Before U.S. Military Killed Them

https://theintercept.com/2025/12/05/boat-strike-survivors-double-tap/
5•belter•16m ago•1 comments

Notes on RLHF Book by Nathan Lambert

https://shubhamg.bearblog.dev/notes-on-rlhf-nathan-lambert-book/
1•shubham13596•17m ago•0 comments

Keep hitting US Big Tech with fines, Europe's Greens tell von der Leyen

https://www.politico.eu/article/keep-hitting-us-big-tech-fines-europe-greens-tell-ursula-von-der-...
3•saubeidl•22m ago•2 comments

Show HN: Sornic – Turn any URL into social media posts for 6 platforms

https://sornic.com
2•digi_wares•23m ago•1 comments

Tree-me: Because Git worktrees shouldn't be a chore

https://haacked.com/archive/2025/11/21/tree-me/
1•wolfwyrd•23m ago•0 comments

Geometric derivation of proton radius matches CODATA within 577 ppm

https://zenodo.org/records/17847771
1•albert_roca•24m ago•1 comments

The new moat in AI isn't models. It's data infrastructure

https://twitter.com/mannamenash/status/1995518780064350607
3•leandrenash•28m ago•2 comments

"Not Always Right": Funny and True Stories

https://notalwaysright.com/
1•ColinWright•29m ago•0 comments

Waymo's self-driving cars passed stopped school buses

https://www.npr.org/2025/12/06/nx-s1-5635614/waymo-school-buses-recall
2•thinkcontext•29m ago•0 comments

The AI productivity struggle is not about AI

https://haugeto.medium.com/the-ai-productivity-struggle-is-not-about-ai-c932c7c76453
1•ingve•30m ago•0 comments

To boost research, states are building their own AI-ready supercomputers

https://www.science.org/content/article/boost-research-states-are-building-their-own-ai-ready-sup...
1•ashishgupta2209•33m ago•0 comments

A Fasting-Style Diet Seems to Result in Dynamic Changes to Human Brains

https://www.sciencealert.com/a-fasting-style-diet-seems-to-result-in-dynamic-changes-to-human-brains
1•makaimc•34m ago•0 comments

Queensland Museum accused of misleading teachers and children about climate

https://www.theguardian.com/australia-news/2025/dec/08/queensland-museum-accused-of-misleading-te...
2•doener•40m ago•1 comments

IntelliJ Scala Plugin 2025.3 Is Out

https://blog.jetbrains.com/scala/2025/12/08/scala-plugin-2025-3-is-out/
1•quapster•40m ago•0 comments

Russia Loses Launch Capability After Accident at Baikonur Cosmodrome

https://www.universetoday.com/articles/severe-accident-destroys-russias-ability-to-launch-astrona...
2•belter•41m ago•0 comments

TurtleStitch – Coded Embroidery

https://www.turtlestitch.org/
1•tosh•42m ago•0 comments

The Internet forgets, but I don't want to

https://alexwlchan.net/2025/social-media-scrapbook/
1•ingve•44m ago•0 comments

Clip of a Tesla Optimus teleoperator taking his headset off

https://bsky.app/profile/jjvincent.bsky.social/post/3m7hps4h2x22r
1•doener•46m ago•0 comments

Google is experimentally replacing news headlines with AI clickbait nonsense

https://www.theverge.com/ai-artificial-intelligence/835839/google-discover-ai-headlines-clickbait...
4•rbanffy•47m ago•3 comments

How to Articulate Yourself Intelligently

https://letters.thedankoe.com/p/how-to-articulate-yourself-intelligently
1•BerislavLopac•48m ago•0 comments

Catala – DSL for deriving algorithms producing automated legal decisions

https://github.com/CatalaLang/catala
2•amai•48m ago•0 comments