Hackernews posts about 32B

1. Granite 4.1: IBM's 8B Model Matching 32B MoE (firethering.com)
   328 points by steveharing1 22 days ago | 209 comments
2. An unknown Sega Saturn project has come to light after 29 years (32bits.substack.com)
   112 points by bbayles 20 days ago | 2 comments
3. The impossible maths of the AI boom (www.ft.com)
   6 points by disgruntledphd2 2 days ago | 1 comments
4. Ramp in talks to hit $40B+ valuation, 6 months after reaching $32B (techcrunch.com)
   1 points by SilverElfin 15 days ago | 1 comments
5. Indexing a year of video locally on a 2021 MacBook with Gemma4-31B (50GB swap) (blog.simbastack.com)
   452 points by asenna 1 day ago | 128 comments
6. Meta to receive $3.3B in tax breaks for its $10B Louisiana data center (fortune.com)
   192 points by logickkk1 7 days ago | 181 comments
7. Only 17% of all 64-bit Integers are products of two 32-bit integers (lemire.me)
   12 points by chmaynard about 11 hours ago | 2 comments
8. Xkcd 3233 – Make It Myself (xkcd.com)
   11 points by arm32 9 days ago | 1 comments
9. My Daughter Died at 32. My Devices Won't Let Me Rest (www.wsj.com)
   10 points by impish9208 22 days ago | 2 comments
10. Show HN: A 0.3B model that redacts PII in all 24 EU languages offline (huggingface.co)
    6 points by mipo57 9 days ago | 3 comments
11. Why do Windows editions on 32-bit x86 systems artificially limit RAM to 4 GB? (devblogs.microsoft.com)
    6 points by supermatou 7 days ago | discuss
12. Startup Raises $1.3B for an A.I. 'Grid' (www.nytimes.com)
    4 points by bookofjoe 7 days ago | 2 comments
13. ByteDance Open-Sources Lance, a 3B Multimodal Model for Images and Video (firethering.com)
    4 points by steveharing1 1 day ago | discuss
14. Microsoft Now Recommends 32 GB RAM as a "No Worries" Upgrade for Windows 11 (www.techpowerup.com)
    4 points by SockThief 21 days ago | discuss
15. Making a 3B Robot Policy Faster (www.hapticlabs.ai)
    4 points by ibero 29 days ago | discuss
16. How Rockstar fit a city into 32 megabytes of PS2 memory [video] (www.youtube.com)
    3 points by newswangerd 7 days ago | discuss
17. Q8 KV cache lets a 30B model fit 100K context on a 24 GB RTX 5090 (buraak.com)
    3 points by bozdemir 16 days ago | discuss
18. LLM 0.32a0 is a major backwards-compatible refactor (simonwillison.net)
    3 points by thebeardisred 22 days ago | discuss
19. Poisoning RAG document corpora: 32 vectors tested, 19 succeeded (corrupted.io)
    3 points by kusuriya 25 days ago | discuss
20. Why is AI trust so much higher in China (87%) than the US (32%)? [pdf] (www.edelman.com)
    2 points by nilen 12 days ago | 9 comments
21. GLiNER2-PII: 0.3B open-source PII model outperforms OpenAI's Privacy Filter (pioneer.ai)
    2 points by neon_share1 8 days ago | 1 comments
22. Plimpton 322 – Babylonian Clay table of triangles 1k years before Pythagoras (en.wikipedia.org)
    2 points by lifeisstillgood 15 days ago | 1 comments
23. MiniCPM-v4.6 challenges larger multimodal models at 1.3B params (firethering.com)
    2 points by steveharing1 9 days ago | discuss
24. Summer of Math Expositions(2021-2025) – 3b1B archive (some.3b1b.co)
    2 points by num42 15 days ago | discuss
25. The end of seats: pricing Netlify for 3B builders (www.netlify.com)
    2 points by seanwilson 18 days ago | discuss
26. Local Harness Benchmark: Pi Coding Agent vs. OpenCode with Qwen3.6 35B A3B (grigio.org)
    2 points by grigio 18 days ago | discuss
27. Building a daily AI news brief in 325 lines of Python (gist.github.com)
    2 points by Alex_Vance 19 days ago | discuss
28. Surviving Black Friday: 329B requests with Falcon (speakerdeck.com)
    2 points by ksec 22 days ago | discuss
29. Stride and prejudice: How a 32-bit overflow corrupted a CUDA kernel (www.ai21.com)
    2 points by kjeetgill 24 days ago | discuss
30. Running Gemma 4 31B on Mac with Ollama (sammyrulez.github.io)
    2 points by sammy_rulez 26 days ago | discuss