frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Gemini 3 Flash: Frontier intelligence built for speed

https://blog.google/products/gemini/gemini-3-flash/
745•meetpateltech•8h ago•382 comments

Gut Bacteria from Amphibians and Reptiles Achieve Complete Tumor Elimination

https://www.jaist.ac.jp/english/whatsnew/press/2025/12/17-1.html
48•Xunxi•2h ago•10 comments

OBS Studio Gets a New Renderer

https://obsproject.com/blog/obs-studio-gets-a-new-renderer
90•aizk•4h ago•22 comments

Coursera to combine with Udemy

https://investor.coursera.com/news/news-details/2025/Coursera-to-Combine-with-Udemy-to-Empower-th...
412•throwaway019254•12h ago•230 comments

Why My Payment Agent Is Named George, Not Stripe-Agent

https://blog.kestrelsnest.social/posts/2025-12-14-why-my-payment-agent-is-named-george-not-stripe...
10•fortyseven•3d ago•4 comments

AWS CEO says replacing junior devs with AI is 'one of the dumbest ideas'

https://www.finalroundai.com/blog/aws-ceo-ai-cannot-replace-junior-developers
723•birdculture•8h ago•413 comments

I got hacked: My Hetzner server started mining Monero

https://blog.jakesaunders.dev/my-server-started-mining-monero-this-morning/
168•jakelsaunders94•4h ago•151 comments

Inside PostHog: SSRF, ClickHouse SQL Escape and Default Postgres Creds to RCE

https://mdisec.com/inside-posthog-how-ssrf-a-clickhouse-sql-escaping-0day-and-default-postgresql-...
70•arwt•4h ago•18 comments

Show HN: High-Performance Wavelet Matrix for Python, Implemented in Rust

https://pypi.org/project/wavelet-matrix/
56•math-hiyoko•5h ago•0 comments

Explaining the Widening Divides in US Midlife Mortality: Is There a Smoking Gun?

https://www.nber.org/papers/w34553
14•bikenaga•1h ago•7 comments

A Safer Container Ecosystem with Docker: Free Docker Hardened Images

https://www.docker.com/blog/docker-hardened-images-for-every-developer/
270•anttiharju•8h ago•55 comments

Tell HN: HN was down

471•uyzstvqs•8h ago•279 comments

Cloudflare Radar 2025 Year in Review

https://radar.cloudflare.com/year-in-review/2025
45•ksec•3h ago•15 comments

TikTok unlawfully tracks your shopping habits – and your use of dating apps

https://noyb.eu/en/tiktok-unlawfully-tracks-your-shopping-habits-and-your-use-dating-apps
28•doener•57m ago•6 comments

Fast Sequence Iteration in Common Lisp

https://world-playground-deceit.net/blog/2025/12/fast-sequence-iteration-in-common-lisp.html
28•BoingBoomTschak•4d ago•5 comments

How SQLite is tested

https://sqlite.org/testing.html
222•whatisabcdefgh•7h ago•55 comments

Zmij: Faster floating point double-to-string conversion

https://vitaut.net/posts/2025/faster-dtoa/
88•fanf2•3d ago•9 comments

Launch HN: Kenobi (YC W22) – Personalize your website for every visitor

28•sarreph•8h ago•49 comments

The Number That Turned Sideways

https://zuriby.github.io/math.github.io/the-number-that-turned-sideways.html
14•tzury•4d ago•4 comments

Speed matters: Why working quickly is more important than it seems

https://jsomers.net/blog/speed-matters
33•bschne•2d ago•19 comments

Flick (YC F25) Is Hiring Founding Engineer to Build Figma for AI Filmmaking

https://www.ycombinator.com/companies/flick/jobs/Tdu6FH6-founding-frontend-engineer
1•rayruiwang•8h ago

Developers can now submit apps to ChatGPT

https://openai.com/index/developers-can-now-submit-apps-to-chatgpt/
63•tananaev•2h ago•45 comments

Pornhub extorted after hackers steal Premium member activity data

https://www.bleepingcomputer.com/news/security/pornhub-extorted-after-hackers-steal-premium-membe...
89•coloneltcb•5h ago•36 comments

I couldn't find a logging library that worked for my library, so I made one

https://hackers.pub/@hongminhee/2025/logtape-fedify-case-study
26•todsacerdoti•5d ago•32 comments

Learning Fortran (2024)

https://uncenter.dev/posts/learning-fortran/
55•lioeters•11h ago•48 comments

Thin desires are eating life

https://www.joanwestenberg.com/thin-desires-are-eating-your-life/
759•mitchbob•2d ago•246 comments

I created a publishing system for step-by-step coding guides in Typst

https://press.knowledge.dev/p/new-150-pages-rust-guide-create-a
27•deniskolodin•4d ago•7 comments

The State of AI Coding Report 2025

https://www.greptile.com/state-of-ai-coding-2025
76•dakshgupta•8h ago•83 comments

Is Mozilla trying hard to kill itself?

https://infosec.press/brunomiguel/is-mozilla-trying-hard-to-kill-itself
820•pabs3•15h ago•723 comments

Show HN: GitForms – Zero-cost contact forms using GitHub Issues as database

https://gitforms-landing.vercel.app/
15•lgreco•6h ago•11 comments
Open in hackernews

Reverse-engineering the RK3588 NPU: Hacking limits to run vision transformers

https://amohan.dev/blog/2025/shard-optimizing-vision-transformers-edge-npu/
51•rcarmo•1d ago

Comments

jauntywundrkind•1d ago
Epic hacker work!

For what it's worth, it seems like there's a bunch of open source NPU work in progress too. There's a layer "TEFLON" for Gallium3D shared by most of these drivers, that TensorFlow can use. Then hardware drivers for Rockchip (via ROCKET driver), and Vivante (with their Etnaviv drivers). It'd be extra interesting now to see how (or if?) they've dealt with the system constraints (small scratchpad size) here. https://www.phoronix.com/news/Gallium3D-Teflon-Merged https://www.phoronix.com/news/Rockchip-NPU-Linux-Mesa https://www.phoronix.com/news/Two-NPU-Accel-Drivers-2026

poad4242•7h ago
> *Thanks! I actually tracked the Teflon/ROCKET driver work closely during my initial research (it was the 'Plan B' in my original proposal if the vendor blobs failed entirely).* >

> *The main reason I stuck with the closed-source `rknn` stack for this specific project was operator support for Transformers. Teflon is getting great at standard CNN ops (Fused ReLU, Convs, etc.), but the SigLIP vision encoder relies on massive Transposes and unbounded GELU activations that currently fall off the 'happy path' in the open stack.*

> *To your point on the system constraints (small scratchpad): I suspect the current open-source drivers would hit the exact same 32KB SRAM wall I found. The hardware simply refuses to tile large matrices automatically. My 'Nano-Tiling' fix was a software-level patch; porting that logic into the Mesa driver itself would probably be the 'Holy Grail' fix here.*

Neywiny•1d ago
This is good work. I would say that there was very little reverse engineering but that's fine. It's interesting seeing some companies look at ARM's Ethos line as holding them back and others as it pulling them forward. I'm not sure if ARM is the best solution, but all these different NPUs feels a bit like the early CPU architecture and compiler days. Hopefully we can make it through unscathed so at least we get better error messages or maybe even compilers that know those kinds of idiosyncracies enough to avoid such things.
kvuj•1d ago
Awesome! Finally putting back "Hacker" in "Hacker News".
doctorpangloss•1d ago
hacker news needs a reprieve from "Problem. The fix? Vibe coding session. Here's the ChatGPT report"
poad4242•7h ago
I understand the frustration with AI-written posts lately, but this was the opposite of that. It took months of hard work and many late nights. While the hardware manual (TRM) is public, it doesn't explain how to handle the strict 4KB memory bank limits. I had to figure out how to shard and tile the model because the hardware won't let you store data across those banks without crashing. It was a long battle with memory constraints to get that 15x speedup.
PunchyHamster•1d ago
we need RISC-V equivalent but for NPUs, it's become a royal mess last few years
Neywiny•1d ago
It's starting. Some designs are moving towards very wide vector length (1k maybe even 2k?) RV-V cores. So less a giant matrix multiplication unit (I think TI has some parts with what they literally call MMUs, great work guys), more a bunch of DSP heavy CPUs. In the age of x86 splitting on AVX-512, it's interesting.
poad4242•7h ago
Hello! Author of the post here, happy to answer questions about the process. I have a draft white paper that details more of the process. Let me know if I should put it up on github or arxiv.