• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
  • HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI

    Chip running Un-0 reportedly uses less than 900 nanojoules to generate an image

    An Un-0 team member claims the chip uses many orders of magnitude less energy per image than conventional machines. A follow-up describes teaching the system to count.

    NR
    SJ
    3 Sources, ,

    TLDR

    In a September 16, 2026 update, an Un-0 team member said hardware had arrived that summer and was running the model released in June. They said the shared images came from actual hardware, claiming energy use below 900 nanojoules per image—many orders of magnitude less than conventional machines. The author also credited AI with tighter research loops, saying the project went from no team in January to tape-out, the design handoff for fabrication, in five months. In a follow-up, they described a sequence generated on the chip and said the system had been taught to count.

    Combined views

    157.8K

    3 Sources, first seen 14d ago

    Combined views

    157.8K

    3 Sources, first seen 14d ago

    710 likes
    14d ago
    first seen 14d ago
    710 likes
    36 comments
    356 saves
    73 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    36 comments
    356 saves
    73 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    3 Sources

    @NaveenGRaoI made some public disclosures about our hardware at the @theallinpod conf this week and wanted to share the progress here. Earlier this summer we got hardware back! The execution was madness…we went from no team in Jan to a tape out in 5 months. AI enabled MUCH tighter loops of research and our execution speed shows the results. This is the first large-scale demonstration of a causal, physical dynamical system to do real compute. 🚀 We released the Un-0 model (see link) in June and it runs on this physical system; the images below come from the actual hardware. What’s more, this chip requires <900 nJ to generate an image; this is many orders of magnitude less energy than conventional machines. https://unconv.ai/blog/introducing-un-0-generating-images-with-coupled-oscillators/
    @FutureJurvetsonThis is bonkers. Unconventional AI’s first analog AI chip (using arrays of coupled resonators) generates an image with a total energy less than 900 nano joules. Analog AI is the next 1000x in power efficiency. Just like our brain.

    3 Sources

    @NaveenGRaoI made some public disclosures about our hardware at the @theallinpod conf this week and wanted to share the progress here. Earlier this summer we got hardware back! The execution was madness…we went from no team in Jan to a tape out in 5 months. AI enabled MUCH tighter loops of research and our execution speed shows the results. This is the first large-scale demonstration of a causal, physical dynamical system to do real compute. 🚀 We released the Un-0 model (see link) in June and it runs on this physical system; the images below come from the actual hardware. What’s more, this chip requires <900 nJ to generate an image; this is many orders of magnitude less energy than conventional machines. https://unconv.ai/blog/introducing-un-0-generating-images-with-coupled-oscillators/
    @FutureJurvetsonThis is bonkers. Unconventional AI’s first analog AI chip (using arrays of coupled resonators) generates an image with a total energy less than 900 nano joules. Analog AI is the next 1000x in power efficiency. Just like our brain.