frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

The Economics of Open-Weight Inference

https://data.ornn.com/publications/the-economics-of-open-weight-inference
11•marinesebastian•2h ago

Comments

jcmontx•36m ago
The problem, IMO, with open-weight models is that you accustom to the capabilities of frontier models too quickly; and downgrading to an open-weight "frontier minus 2" or "frontier minus 3" model is often painful, since they feel way less useful than their newer closed-weights counterpart. To be honest, I don't know any companies using OW models at a large scale for their operations (agents or chat assistants).
augment_me•33m ago
I think an interesting point is that hardware as of today still has no utility value after its reported lifetime has elapsed, which prevents neolabs and smaller labs from getting older HW clusters as the banks are not willing to give out loans against them. There is no agreed upon pricing for "expired" A100 clusters or similar.

This is clearly not true, and we are starting to see compute markets, but only for rental prices/H, not for the hardware itself. I feel like there is some artificial moat being built here to stimulate sales of new hardware, because an H100 at 1/16th the price will have comparable dollar/FLOP as Vera Rubin.

linuxftw•23m ago
Depends on the workload. H100 will never have the network performance of Vera Rubin. There's also token per watt, newer systems will beat the older systems.
kurthr•11m ago
It's not clear how much of the latest chips have even made it on-line yet.

The claims of many GW of installed training/inference have come under scrutiny lately. The first VeraRubins aren't even there yet, so it's all GB300 NVL72s as the peak performers and probably <<1GW of those so far. Even xAI Colossus is mostly H200s and B200s.

Electricity costs are also a huge differentiator. When drawing 100kW the difference between >50cents and <10cents per kWh is pretty big! One is almost $0.5M and the other is less than $100k.

simianwords•16m ago
Then why do people still believe that OpenAI any Anthropic have negative margins
hungryhobbit•3m ago
Dark grey text on a black background: it's like they are trying not to let anyone read their article!

OpenAI GPT–6 Astra breaks Enigma message that has resisted solution since 2005

https://www.cryptocellar.org/bgac/the-mvueh-break.html
263•sohkamyung•2h ago•251 comments

Claude Opus 5.5

https://www.anthropic.com/claude-opus-5-5
49•throwaway371647•9m ago•28 comments

16-bit Intel 8088 chip

https://allpoetry.com/16-bit-Intel-8088-chip
28•rbanffy•25m ago•2 comments

OpenAI is well positioned to fast-follow Jev

https://arcturus-labs.com/blog/2026/09/21/will-openai-eat-jevs-lunch/
107•JohnBerryman•1h ago•83 comments

Apple has added persistent 'ads' to iOS, and it's driving users crazy

https://www.techradar.com/phones/iphone/i-wish-apple-would-just-stop-that-crap-apple-has-added-pe...
290•MC995•2h ago•201 comments

Writing Rust code that's fast by asking agents to make the code faster

https://minimaxir.com/2026/09/agentic-iteration/
21•mooreds•58m ago•4 comments

Show HN: Drop – a rootless Linux sandbox with gVisor support

https://droprun.sh/
84•mixedbit•2h ago•24 comments

Solitaire Alone Together

https://solitairealonetogether.com/
42•eieio•18h ago•4 comments

Can gzip be a language model?

https://nathan.rs/posts/gzip-lm/
317•networked•10h ago•119 comments

MiMo v2.6

https://mimo.xiaomi.com/mimo-v2-6
1032•volf_•20h ago•458 comments

I asked Meta’s Muse for its filesystem and it sent me 6.8GB

https://mouse.dev/blog/muse-runtime-export/
136•Aeroi•1h ago•73 comments

Spymarks, Not Watermarks

https://brand.io/article/spymarks/
608•possibilistic•17h ago•154 comments

Muse, Meta's extraordinarily privileged AI assistant, has a serious 0-day

https://arstechnica.com/security/2026/09/muse-metas-extraordinarily-privileged-ai-assistant-has-a...
57•pavel_lishin•2h ago•23 comments

Xbox continues its "reset" with dramatic restructuring

https://arstechnica.com/gaming/2026/09/shuffling-the-deck-chairs-xbox-continues-its-reset-with-dr...
13•dgrin91•26m ago•1 comments

Jev – A curation of Jev demos on X, tools, skills, and integrations

https://github.com/Amal-David/awesome-jev
72•frostbyte7•1h ago•20 comments

Line Scan Photos from MUNI Heritage Weekend in San Francisco

https://daniel.lawrence.lu/blog/2026-09-20-muni-heritage-weekend/
94•plun9•1d ago•16 comments

Teleoperated Humans

https://www.jefftk.com/p/teleoperated-humans
25•luu•2d ago•19 comments

Attention is all you have

https://alicegg.tech/2026/09/21/attention
994•zer0tonin•1d ago•297 comments

Vacate a Drone Restriction That Criminalized Recording Immigration Agents

https://www.eff.org/deeplinks/2026/09/dc-circuit-must-vacate-drone-flight-restriction-criminalize...
28•hn_acker•1h ago•7 comments

Transformers Explained Visually

https://poloclub.github.io/transformer-explainer/
559•aray07•20h ago•84 comments

Relativistic Raytracing

https://publish.obsidian.md/h1m3/Articles/Relativistic+raytracing
3•vismit2000•1d ago•0 comments

The Economics of Open-Weight Inference

https://data.ornn.com/publications/the-economics-of-open-weight-inference
11•marinesebastian•2h ago•6 comments

What Sun got wrong

https://bcantrill.dtrace.org/2026/09/20/what-sun-got-wrong/
659•chmaynard•1d ago•379 comments

I said no and Apple said yes

https://dbushell.com/2026/09/22/apple-intelligence/
629•thatslast•8h ago•503 comments

A font that reads what you wrote

https://rohanadwankar.github.io/posts/semfont.html
69•RohanAdwankar•2d ago•37 comments

I don't want to read what you didn't write

https://blog.colinbreck.com/i-dont-want-to-read-what-you-didnt-write/
920•mooreds•18h ago•386 comments

AMD's random number generator can't generate a 0?

https://board.flatassembler.net/topic.php?t=24261
187•BruceEel•7h ago•145 comments

MiMo-v2.6-Pro: Intelligence, Performance and Price Analysis

https://artificialanalysis.ai/models/mimo-v2-6-pro
133•theanonymousone•12h ago•60 comments

AI coding has made CI a bottleneck, so we reworked ours to keep up

https://linear.app/now/ci-bottleneck-reworked
298•julian_digital•21h ago•365 comments

Engineering Memory: On learning to memorize first 100 digits of pi (2024)

https://gregorygundersen.com/blog/2024/12/21/engineering-memory/
51•fuzzythinker•1d ago•32 comments