I’m looking to get a GPU to start experimenting with different AI models.
I am curious, however, under what conditions is it more cost-effective to attempt to run your AI on your own hardware vs paying a third party?
I’m looking to get a GPU to start experimenting with different AI models.
I am curious, however, under what conditions is it more cost-effective to attempt to run your AI on your own hardware vs paying a third party?
Tim Dettmers has a nice writeup exploring this question (and a variety of related ones).
I can’t post links at the moment, but try searching the web for “tim dettmers gpu guide”
and then within that page, search for: “When is it better to use the cloud vs a dedicated GPU desktop/server?”
This shows some good ones you can run locally. I ws interested to hear you can download Facebook’s 20 million dollar pre-trained AI for yourself. lol
This topic was automatically closed 273 days after the last reply. New replies are no longer allowed.