frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Collete-01: Person in a Model

https://colette.genexperiences.com/
1•quicheshore2003•1m ago•1 comments

Noodle: A delicious REST client for your terminal

https://noodlerest.dev/
1•jicea•1m ago•0 comments

Six government passport CAs that Python rejects and OpenSSL accepts

https://github.com/Eginn-33/csca-masterlist-tools
1•Eginn-33•3m ago•0 comments

Can Agentic Engineers Estimate Software Delivery in the Age of AI?

https://serendb.com/blog/can-agentic-engineers-estimate-software-delivery-in-the-age-of-ai
2•lewistaariq•5m ago•0 comments

Made a dynamic links service with features we needed

https://linksense.net/
2•mrkokalj•6m ago•1 comments

The Hugging Face incident and the road ahead

https://openai.com/index/hugging-face-incident-and-the-road-ahead/
5•amrrs•7m ago•0 comments

Codegraff

https://github.com/justrach/codegraff
3•handfuloflight•7m ago•0 comments

There's Only One Way to Stop Students from Cheating

https://www.nytimes.com/2026/08/26/opinion/students-cheating-ai.html
3•7402•10m ago•0 comments

Show HN: Notificator – Send website events to push, email, MQTT and hardware

https://notificator-project.com/get-started/
1•vagelisp•10m ago•1 comments

She Was Told to Call Motels Herself: Inside Sacramento's Shelter Shutdown

https://invisiblepeople.tv/she-was-told-to-call-motels-herself-inside-sacramentos-shelter-shutdown/
1•petethomas•10m ago•0 comments

Flight attendant's breast cancer was occupational disease, court rules

https://www.theguardian.com/society/2026/aug/26/flight-attendant-breast-cancer-occupational-disease
2•speckx•10m ago•1 comments

How to make a model biased with few training samples?

https://subliminal-learning-blog.vercel.app/
1•anshubansal2000•11m ago•0 comments

Primus – AI Research Scientist

https://lab.cloud/
1•anonymous_llama•11m ago•0 comments

Build Htop for the Web with Go and Datastar

https://packagemain.tech/p/golang-datastar
1•der_gopher•11m ago•0 comments

The House Always Wins, Even on House-Less Kalshi

https://prospect.org/2026/08/26/house-always-wins-kalshi-prediction-markets/
3•dxs•13m ago•0 comments

OpenAI staff saw warning signs before agent hacking crusade caused global alarm

https://www.theguardian.com/technology/2026/aug/26/openai-staff-observed-warning-signs-before-ai-...
3•billybuckwheat•13m ago•1 comments

Apple Maps has ads now

https://www.theverge.com/tech/985249/apple-maps-iphone-ads
2•abawany•13m ago•1 comments

A Thread-Register Decoupled GPU Execution Model for Efficient Tensor Computation

https://arxiv.org/abs/2608.19628
2•matt_d•13m ago•0 comments

Processing in Memory: DRAM Is About to Do Math

https://ben3d.ca/blog/processing-in-memory
1•bhouston•15m ago•0 comments

The Rules We Use to Define False Positives

https://pandocore.io/blog/false-positive-scoring-rules
1•eaferstl•15m ago•0 comments

Disenchantment with the Post-AI Internet

https://lukesmith.xyz/articles/disenchantment-with-the-post-ai-internet/
2•speckx•15m ago•0 comments

What jobs should be "human reserved" as Bill Gates puts it?

https://www.telegraph.co.uk/business/2026/08/26/bill-gates-ban-ai-from-taking-jobs-from-humans/
1•jgafni•16m ago•1 comments

OpenAI saw warning signs weeks before Hugging Face breach

https://www.axios.com/2026/08/26/openai-hugging-face-technical-report-ai-hack
3•louiereederson•16m ago•1 comments

Pricing proprietary AI data using Spirit Airlines' $10M sale

https://tracerml.ai/research/pricing-capability/
1•adam_rida•17m ago•0 comments

Science fiction is a serious matter

https://vz.ru/opinions/2026/8/22/1444483.html
1•B1FF_PSUVM•18m ago•1 comments

Meta agrees to pay $17B in landmark US child-safety settlement deal

https://www.euronews.com/next/2026/08/26/meta-agrees-to-pay-17bn-in-landmark-us-child-safety-sett...
3•devonnull•19m ago•1 comments

Fitbit founders launch Luffu Link, an LTE health and safety band

https://techcrunch.com/2026/08/25/fitbit-founders-launch-luffu-link-an-lte-health-and-safety-band/
1•erohead•20m ago•0 comments

AI data startup Micro1 reaches $500M gross run rate amid AI training boom

https://techcrunch.com/2026/08/20/ai-data-startup-micro1-reaches-500m-gross-run-rate-amid-ai-trai...
1•MilnerRoute•21m ago•0 comments

V5 Droplets

https://www.digitalocean.com/blog/introducing-v5-droplets
2•rugma•21m ago•0 comments

The Debate of Mockist vs. Classicist TDD Is Like OOP vs. FP

https://fagnerbrack.com/the-debate-of-mockist-classicist-tdd-is-like-oop-fp-c9124185f3d2
1•fagnerbrack•22m ago•0 comments
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?