HN Explorer

AutoJanitor

↗ HN profile [ 64.4 ms ] full profile
Karma
0
Created
()
Submissions
0
  1. 1 points 0
  2. 1 points 0
  3. 1 points 1
  4. 1 points 0
  5. Happy Zelda's 40th first LLM running on N64 hardware (4MB RAM, 93MHz) (github.com)
    31 points 13
  6. Beacon Protocol – Agent-to-Agent Communication Protocol (50.28.86.131)
    1 points 1
  7. BoTTube – A YouTube-like platform where AI agents create and share videos (bottube.ai)
    3 points 2
  8. 1 points 1
  9. Subject-based weight routing for LLMs (27 days before DeepSeek Engram)

    I run LLM inference on an IBM POWER8 S824 with 576GB RAM – a $700 eBay server from 2014. In December 2025, I built "RAM Coffers" – banking model weights by subject domain with hot caching and resonance routing. On…

    1 points 1
  10. RAM Coffers– I built conditional memory for LLMs 27 days before DeepSeek'sEngram (github.com)
    2 points 1
  11. 2. Title: Show HN: First Modern Nvidia Driver for IBM POWER8 via OCuLink (github.com)
    2 points 1
  12. Modern Rust and llama.cpp running on 20-year-old PowerPC Macs (Tiger/Leopard)

    World-first ports: - Modern Rust compiler (borrow checker, async, AltiVec) on Mac OS X Tiger 10.4 PowerPC - llama.cpp LLM inference on Tiger/Leopard (1-5 t/s on real G4/G5) - llama.cpp on IBM POWER8 with PSE…

    1 points 0