fp.

Show HN: Superjson – Simple, beautiful JSON explorer

https://superjson.dev/

1•nihalwashere•2m ago•0 comments

Show HN: Minimal Pomodoro timer for macOS (1.7MB, now with keyboard shortcuts)

https://apps.apple.com/us/app/pomodoro-timer-lite/id6748662476?mt=12

1•happylaodu•2m ago•0 comments

FAA Lifts Closure at El Paso Airport

https://www.nytimes.com/live/2026/02/11/us/faa-el-paso-flights-airport

1•joekrill•3m ago•0 comments

An AI-generated pull request that makes sense

https://nicolaiarocci.com/an-ai-generated-pull-request-that-actually-makes-sense/

1•speckx•3m ago•0 comments

Deploying Rust to Production Checklist

https://kerkour.com/rust-production-checklist

3•randomint64•4m ago•0 comments

Show HN: Triclock – A Triangular Clock

https://triclock.franzai.com/

1•franze•4m ago•0 comments

Show HN: Deeploy v0.3.0 – terminal-first VPS app deployment tool

https://deeploy.sh

1•axadrn•5m ago•0 comments

Show HN: Eryx, a fast WASM-based Python sandbox with native extension support

https://github.com/eryx-org/eryx

1•sd2k•6m ago•0 comments

Don't Go Monolithic; the Enterprise Agent Stack Is Stratifying

https://philippdubach.com/posts/dont-go-monolithic-the-agent-stack-is-stratifying/

1•7777777phil•7m ago•0 comments

Ejabberd 26.02 / ProcessOne – Erlang Jabber/XMPP/Matrix Server – Communication

https://www.process-one.net/blog/ejabberd-26-02/

1•neustradamus•8m ago•0 comments

Sony to exit Blu-ray disc recorder market amid rise of streaming services

https://english.kyodonews.net/articles/-/70241

1•thunderbong•8m ago•0 comments

Ask HN: What cool things are people building with OpenClaw?

1•gregagi•8m ago•1 comments

Show HN: On-Call Health – Spot signs of overload in incident responders

1•jjtang1•10m ago•0 comments

Is anyone using AI for good?

https://stackoverflow.blog/2026/02/11/is-anyone-using-ai-for-good/

1•HieronymusBosch•11m ago•0 comments

Show HN: SNKV: 49K vs. 43K ops/SEC (70% Reads), 2MB vs. 20MB RAM vs. RocksDB

https://github.com/hash-anu/snkv

1•hashmakjsn•13m ago•0 comments

Architecting on Cloudflare - Book by Jamie Lord

https://architectingoncloudflare.com/

2•ianwootten•15m ago•1 comments

I built a free When2Meet alternative that works on mobile

https://groupmeetingscheduler.com/

2•hasitha_ch•16m ago•1 comments

Show HN: Spheni – A Vector Search Engine in C++ built from scratch, for Python

https://github.com/datavorous/spheni

2•datavorous_•16m ago•0 comments

Can India be a "third way" AI alternative to the U.S. and China?

https://restofworld.org/2026/india-ai-summit-third-way-global-south-big-tech/

1•NDAjam•17m ago•1 comments

The Epstein Tapes

https://www.thefp.com/p/watch-the-jeffrey-epstein-tapes-part

3•belter•19m ago•0 comments

How I Developed Netlify Capsules AR Experience with Nuxt 4 and Three JS

https://www.leemartin.com/netlify-capsules/

2•leemartin•19m ago•1 comments

Running Pydantic's Monty Rust Sandboxed Python Subset in WebAssembly

https://simonwillison.net/2026/Feb/6/pydantic-monty/

1•Einenlum•20m ago•0 comments

OLIX: Compute Manifesto

https://olix.com/blog/compute-manifesto

1•stevanboljevic•21m ago•0 comments

Show HN: Onlybots.cam

https://onlybots.cam

1•m0rtyn•21m ago•1 comments

Show HN: Bulk Localizations and PPP Pricing for AppStore

https://appstorelocalization.com

1•BigBalli•21m ago•1 comments

Stoat is an open-source, user-first chat platform

https://github.com/stoatchat

1•doener•21m ago•0 comments

Originators of Information

https://csmeyer.substack.com/p/originators-of-information

1•csmeyer•22m ago•0 comments

Half of xAI's founders left the company

https://techcrunch.com/2026/02/10/nearly-half-of-xais-founding-team-has-now-left-the-company/

2•Einenlum•22m ago•0 comments

Building a semantic search engine in ±250 lines of Python

https://bart.degoe.de/building-a-semantic-search-engine-in-250-lines-of-python/

4•bartdegoede•24m ago•0 comments

First to Be Second: Why Enduring AI Companies Will Be Started After This Wave

https://babin.posthaven.com/first-to-be-second-why-enduring-ai-companies-will-be-started-after-th...

1•nibab•24m ago•0 comments

Open in hackernews

Building a Zero-Dependency secp256k1 CUDA Engine from Scratch (2.5B ops/SEC)

https://github.com/shrec/UltrafastSecp256k1

1•shrecshrec•1h ago

Comments

shrecshrec•1h ago

I implemented a full secp256k1 engine from scratch in C++ and CUDA with zero external dependencies (no GMP, no OpenSSL).

The goal was to explore performance limits of:

Jacobian mixed-add

Batch inversion using Montgomery’s trick

Large-scale scalar stepping

GPU memory coalescing strategies

On RTX 5060 I’m getting ~2.5B mixed-add operations/sec.

Key design decisions:

Little-endian limb layout for hardware efficiency

Big-endian only for visualization

Deterministic memory layout

No dynamic allocation in hot paths

Would love feedback from people working on ECC or GPU math.