The Clouds Can Do Better Than You Can

There's a capacity problem where you just can't get enough capacity for certain things that you're doing. And to ultimately, at the od of the day, for like training large work loads, it's like a price to performance ratio. So on one hand, you can save like, a lot of money, and certain again, bespoke work loads there. However, there was an interesting conversation that i actually had earlier to day where a company was attempting to run some work loads on pram as relates to gpus. They were just seeing that they couldn't replicate the performance, actually, of some of these large public clob providers. The clouds can do t better than you can