Hackernews posts about 480B

  1. Show HN: Running 104GB Qwen3.8-Flash-Next on 48GB Mac with at ~12 tok/s (github.com)
    240 points by carloslfu 19 days ago | 118 comments
  2. Intel Crescent Island GPU Flexes 32 Xe3P Cores, 480GB LPDDR5X for Agentic AI (hothardware.com)
    7 points by rbanffy 25 days ago | discuss
  3. Intel Crescent Island 160GB to 480GB LPDDR5X AI GPU at Hot Chips 2026 (www.servethehome.com)
    5 points by rbanffy 26 days ago | discuss
  4. Show HN: We tried to recover blurred, pixelated and redacted text (480 cases) (datablur.app)
    4 points by legitimate_key 20 days ago | discuss
  5. Show HN: Skatanica – Browse and share skate clips by location (skatanica.com)
    13 points by ralusek 18 days ago | 3 comments
  6. Show HN: Game to see if you're over-/under-confident in your own knowledge (est.alejo.ch)
    3 points by afc 26 days ago | discuss
  7. Show HN: Neither HTML nor Markdown is enough: a way out of the AI doc dilemma
    1 points by xiongjy2104 26 days ago | 2 comments
  8. Ask HN: What is the best local model that runs on your Mac at decent speed?
    1 points by carloslfu 4 days ago | discuss
  9. $40T Debt, 6.7% Mortgages and $5 Diesel (www.ft.com)
    19 points by tcp_handshaker 29 days ago | 9 comments
  10. OpenAI Claims Latest Model Astra Could Be Considered AGI (www.ft.com)
    3 points by znq 17 days ago | 1 comments
  11. Why performant code matters (but gets widely ignored) [video] (www.youtube.com)
    2 points by HodaSalim 25 days ago | discuss
  12. Norway's Oil Fund Proposes Selling Roughly $80B in U.S. Treasurys (www.wsj.com)
    48 points by toomuchtodo 16 days ago | 17 comments
  13. Nvidia's $200B 'balance sheet-as-a-service' (www.ft.com)
    5 points by JumpCrisscross 25 days ago | 1 comments
  14. I Made a Real Fly Brain Play Pong. It Didn't Learn. That's the Interesting Part (jonatasperaza.medium.com)
    1 points by sebg 10 days ago | discuss
  15. Qwen3 Coder 480B is Live on Cerebras (www.cerebras.ai)
    47 points by retreatguru about 1 year ago | 10 comments
  16. Run Qwen3-Coder-480B-A35B Locally with Unsloth Dynamic Quants (docs.unsloth.ai)
    8 points by handfuloflight about 1 year ago | discuss
  17. Show HN: Qwen3-Coder API – 480B open-source code LLM (www.netmind.ai)
    3 points by elricwan about 1 year ago | 1 comments
  18. Yann LeCun raises $1B to build AI that understands the physical world (www.wired.com)
    612 points by helloplanets 6 months ago | 507 comments
  19. Show HN: The Frontier, Tracking the LLM Pareto Frontier (the-frontier.app)
    3 points by foke82 7 months ago | discuss
  20. Show HN: Running Claude Code using any LLMs? Here's an open-source tool for it (github.com)
    3 points by tensorblock about 1 year ago | discuss
  21. The Rule of Four vs. RFC 3021: Temporal Conflicts in LLM Weights
    2 points by mehrdadrad 7 months ago | discuss
  22. Yann LeCun's AI startup raises $1B in Europe's largest ever seed round (www.ft.com)
    418 points by ottomengis 6 months ago | 2 comments
  23. TikTok owner ByteDance valued at $480B after bidding war (www.thetimes.com)
    6 points by JumpCrisscross 10 months ago | discuss
  24. Thinking Machines draws from Chinese rivals in debut AI model (www.ft.com)
    5 points by tosh 2 months ago | discuss
  25. Anthropic invests $50B in American AI infrastructure (www.diffusemind.com)
    3 points by karmsingh 10 months ago | 2 comments
  26. iPhone 17 Pro Demonstrated Running a 400B LLM (twitter.com)
    713 points by anemll 6 months ago | 326 comments
  27. LongCat-2.0, a large-scale MoE model with 1.6T total and 48B Active (longcat.chat)
    281 points by benjiro29 3 months ago | 88 comments
  28. Trinity large: An open 400B sparse MoE model (www.arcee.ai)
    231 points by linolevan 8 months ago | 82 comments
  29. Cerebras achieves 2,500T/s on Llama 4 Maverick (400B) (www.cerebras.ai)
    93 points by ByteAtATime over 1 year ago | 93 comments
  30. Starlink satellites being lowered from 550 km to 480 km (twitter.com)
    62 points by wmf 9 months ago | 41 comments
  31. Creating a 48GB Nvidia RTX 4090 GPU – Brother Zhang's Repair Shop (Ft. 张哥) [video] (www.youtube.com)
    56 points by adityaathalye 8 months ago | 12 comments
  32. $142 upgrade kit and spare modules turn Nvidia RTX 4090 24GB to 48GB AI card (www.tomshardware.com)
    49 points by RachelF about 1 year ago | 11 comments
  33. Microsoft Fires 4800 as Xbox Division Undergoes Major Reorganization (www.ibtimes.com)
    41 points by justacrow 3 months ago | 11 comments
  34. China launches HDMI and DisplayPort alternative – GPMI up to 192 Gbps, 480W (www.tomshardware.com)
    33 points by ksec over 1 year ago | 9 comments
  35. Linux kernel 7.0 abandons the 28-year-old Intel 440BX chipset (www.tomshardware.com)
    17 points by teleforce 7 months ago | 5 comments
  36. GPMI set to deliver up to 192Gbps and 480W through a single USB cable (videocardz.com)
    13 points by croes over 1 year ago | 9 comments
  37. TagEnergy to build France's largest battery, 240MW/480MWh (www.ess-news.com)
    12 points by gnabgib almost 2 years ago | 17 comments
  38. Custom GeForce RTX 4090 48GB (videocardz.com)
    10 points by doener over 1 year ago | discuss
  39. We analyzed 5,480 hospital cost reports. Supply spending varies 3-7x (andrewrexroad.substack.com)
    8 points by rexroad 6 months ago | 1 comments
  40. V-Color 96GB DDR5-5600 48GB X2 ECC Udimm Kit Quick Look (www.servethehome.com)
    8 points by teleforce over 1 year ago | 1 comments
  41. Intel Announces Arc Pro B60 24GB, Dual B60 48GB (videocardz.com)
    7 points by Topfi over 1 year ago | 3 comments
  42. Nex N2 Pro: Frontier agentic performance at 400B (huggingface.co)
    7 points by himata4113 3 months ago | discuss
  43. China Develops HDMI Alternative: 192 Gbps Speeds and 480 W Power Delivery (www.techpowerup.com)
    7 points by speckx over 1 year ago | discuss
  44. Show HN: A tiny (480 LOC) AI coding assistant for your shell (github.com)
    5 points by n_kr about 1 year ago | 2 comments