4 ms·

Sure, but we could have Hetzners and OVHs who just provide the compute for whatever model we want to run.

• medi8r 8mo ago
  Checked the DDR5 price lately?
  • layer8 8mo ago
    I didn’t claim that it would be cheap. But I’d rather see the real cost of SOTA LLM use exposed. On the other hand, reportedly SOTA LLM inference is profitable nowadays, so it can’t be that expensive.