We often discuss the total amount of GPU compute companies are installing. These clusters are getting huge, with XAI in the lead for the moment. Tesla is able to rival, it not top them with their cluster in Austin. Meta is not far behind, looking to train Llama on even more compute than before.
In this video I dive into what makes Tesla compute different. It isnt the machines as much as what is running on them.
▶️ 3Speak