alt Hacker NewsWould the AI “GW-scale” clusters be able to run the Top500 benchmarks meaningfully? And what might be the outcome?
Replies
wmf • yesterday at 9:21 PM
Yes, they should score well on Linpack as long as they use Ozaki emulation.
➕ show 1 reply
Would the AI “GW-scale” clusters be able to run the Top500 benchmarks meaningfully? And what might be the outcome?
Yes, they should score well on Linpack as long as they use Ozaki emulation.
> And what might be the outcome?
DoE compute budgets are ~10B USD across labs. AI training is a trillion-dollar workload. Different league.