Chinese company DeepSeek announced its new R1 model on January 20. They released a paper on how R1 was trained on January 22. Over the weekend, the DeepSeek app became the number-one free download …
ah, I stand corrected! the figures I was looking at previously were for doing it at acceptable speeds in a data center.
can you imagine the intensity of the RGB in the boy genius Prompt Engineer’s new $6000 custom top end gaming PC with server components? maybe they’ll have the LLM slowly plagiarize them a Python script that turns on more RGB when the GPU’s under load.
so you can run the good version at home! this thread tells how to build a workstation for it.
tl;dr 768GB RAM.
with that, you can run the largest deepseek model, or even open a tab in chrome
apparently it’s not very fast, but it does in fact do the stuff
ah, I stand corrected! the figures I was looking at previously were for doing it at acceptable speeds in a data center.
can you imagine the intensity of the RGB in the boy genius Prompt Engineer’s new $6000 custom top end gaming PC with server components? maybe they’ll have the LLM slowly plagiarize them a Python script that turns on more RGB when the GPU’s under load.