How related is this to AlphaFold and the release of the giant dataset of protein folds? Perhaps there is less of a need for finding the folds and more for understanding them? I heard someone in the know say it will take decades to go through the AlphaFold treasure trove
Currently it has around 3 pflops. That's even lower than what it had in 2018 when it had around 71 pflops and much much lower than its peak during the COVID pandemic when it had 2 exaflops.