Re:How Taalas "prints" LLM onto a chip?
by cpldcpu on Sunday February 22, 2026 @11:37AM
(#47112417)
They mentioned that they using strong quantization (iirc 3bit) and that the model was degradeted from that. Also, they don't have to use transistors to store the bits.
[ Parent
| on HN ]
I think they are talking about the transistors that apply the weights to the inputs.
Re:How Taalas "prints" LLM onto a chip?
by mirekrusin on Sunday February 22, 2026 @03:24PM
(#47114297)
gpt-oss is fp4 - they're saying they'll next try mid size one, I'm guessing gpt-oss-20b then large one, i'm guessing gpt-oss-120b as their hardware is fp4 friendly