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•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?

Training within 10 hours of bedtime measurably hurts recovery

https://tryterra.co/research/best-time-to-workout-for-sleep
1•kyriakosel•3m ago•0 comments

Company Was an American Success Story. Until MAHA Influencers Sank It

https://www.thefp.com/p/this-company-was-an-american-success-until-maha-sank-it
1•rmason•3m ago•0 comments

Trump Order Would Push Banks to Check Clients' Citizenship Status

https://www.wsj.com/finance/regulation/trump-order-would-push-banks-to-check-clients-citizenship-...
1•petethomas•8m ago•0 comments

You Can't Park at 0.1c

https://samirvarma.substack.com/p/you-cant-quietly-park-at-01c
1•qsi•9m ago•0 comments

What are account recovery options with FusionAuth?

https://fusionauth.io/community/forum/topic/3133/what-are-account-recovery-options-with-fusionauth
1•mooreds•14m ago•0 comments

Why "Fast AI" and "Safe AI" Were Never in Conflict

https://www.enkryptai.com/blog/why-fast-ai-and-safe-ai-were-never-actually-in-conflict
1•mooreds•15m ago•0 comments

Etymonline: An Online Etymology Dictionary

https://www.etymonline.com
1•airstrike•16m ago•0 comments

Debunking the job-hopping myth: A Data-Driven Look at Tenure and Turnover [pdf]

https://www.nirsonline.org/wp-content/uploads/2025/09/25-Debunking-Job-Hopping-Myth_FINAL.pdf
2•gmays•18m ago•0 comments

Enterprise AI: Mystery Meat, Kill Zones, Cognitive Surrender, Vibe Bombs

https://kyield.com/insights/newsletter/2026/05/vibe-bombs-cognitive-surrender.html
1•mooreds•18m ago•0 comments

Apple Taps Virtual Avatar Firm Animato's Expertise and Intellectual Property

https://www.macrumors.com/2026/05/19/apple-acquires-animato/
1•mgh2•20m ago•0 comments

Cyoda-Go: The Enterprise Edbms

https://github.com/Cyoda-platform/cyoda-go
1•petethomas•24m ago•0 comments

Alternatives to HN for "tech outside of AI" discussion?

2•summonerOS•25m ago•0 comments

Customizing an LLM for Enterprise Software Engineering

https://arxiv.org/abs/2605.16517
1•azhenley•25m ago•0 comments

Zephex is hosted MCP that gives AI coding editors persistent project context

https://zephex.dev
1•zephex•29m ago•0 comments

Repugnant Economics

https://marginalrevolution.com/marginalrevolution/2026/05/repugnant-economics.html
3•paulpauper•33m ago•0 comments

Show HN: SafeRun – Replay debugging and inline prevention for AI agents

1•Tidianez•33m ago•0 comments

Old Space, New Space: A Commercial Revolution in Innovation?

https://www.nber.org/papers/w35212
1•paulpauper•33m ago•0 comments

Australia and Pax Silica: The Quiet Foundations of a New Western Order?

https://sldinfo.com/2026/05/australia-and-pax-silica-the-quiet-foundations-of-a-new-western-order/
1•Gaishan•34m ago•0 comments

Sen. Cassidy casts deciding vote in legislation to end Iran war [video][40s]

https://www.youtube.com/watch?v=S-h8oM7kGEE
2•Bender•35m ago•0 comments

Vacuum flux has memory too [4]

1•sargstuff•35m ago•0 comments

Meta Begins 8k Global Job Cuts in Asian Hub of Singapore

https://finance.yahoo.com/sectors/technology/articles/meta-begins-8-000-global-004153394.html
3•doppp•38m ago•0 comments

Copyright Office Rejected My Attempt to Copyright a Tweet

https://www.techdirt.com/2014/08/04/copyright-office-rejected-my-attempt-to-copyright-tweet/
2•danhite•43m ago•1 comments

Ember: 365-day audited record of AI models vs. Polymarket, scored by Brier

https://emberfyi.com/
1•emberfyi•44m ago•0 comments

AI video editing is blowing my mind

https://aivideoediting.io/
1•pekingzcc•49m ago•0 comments

Architect of the UK Online Safety Act Calls for Its Complete Repeal

https://prestonbyrne.com/2026/05/19/architect-of-the-uk-online-safety-act-calls-for-its-complete-...
5•iamnothere•50m ago•0 comments

Rendezvous: A serverless, Zoom-like video conferencing web app

https://github.com/predatorray/rendezvous
2•zetaplusae•56m ago•0 comments

Generalization Dynamics of LM Pre-Training

https://jiaxin-wen.github.io/blog/generalization-dynamics
1•gmays•57m ago•0 comments

Iteration

https://blog.viktomas.com/posts/iteration/
1•luca-sctr•58m ago•0 comments

GPU telemetry anomaly: 146W idle draw on A100 (white paper)

https://github.com/mikebains41-debug/ai-gpu-energy-optimizer-/blob/main/WHITEPAPER.md
1•mikebains•1h ago•0 comments

Who Wins the Future: Chips vs. Frontier LLMs

https://medium.com/@vektormemory/who-wins-the-future-chips-vs-frontier-llms-1e8e0ca42641
2•vektormemory•1h ago•1 comments