https://localrig.com
I found this site by chance through a Google search, and it's very well organized.
You can see how many tokens per second you get when running multiple models on your graphics card GPU,
and whether you can run the corresponding LLM model on your graphics card.
It's great to be able to know the speed when purchasing a graphics card to run LLM models initially.
Mac M4 and M5 can handle large capacities, but the speed is surprisingly slow.
I bought an Intel A770 16G for a cheap price, and the speed was surprisingly good, which was comforting. :)
▶ Original source: https://localrig.com
