fcalva

@carrylex @kalkulat If you zoom in, you will see that it is in fact a real photo. It’s just someone who thought putting a CPU upside down was a good idea.

@Treczoks @flemtone Thing is, the final LLM inference is usually done at reduced precision. 8-16 bits usually, but even 4bits or lower with different layers of varying precision.

@TheBlackLounge @kalkulat LLM inference is definitely theoretically possible on analog chips. They just may not scale :v