Hacker News new | ask | show | jobs
by ReptileMan 509 days ago
And then some chineese startup create an amazing compiler that takes cuda and moves it to X (AMD, Intel, Asic) and we are back at square one.

So far it seems that the best investment is in RAM producers. Unlike compute the ram requirements seem to be stubborn.
1 comments
01100011 509 days ago
Don't forget that "CUDA" involves more than language constructs and programming paradigms.

With NVDA, you get tools to deploy at scale, maximize utilization, debug errors and perf issues, share HW between workflows, etc. These things are not cheap to develop.

link
Symmetry 509 days ago
It might not be cheap to develop them but if you can save $10B in hardware costs by doing so you're probably looking at positive ROI.
link
01100011 509 days ago
Yeah, I mean, 9 women can make a baby in a month so why not?

Oh wait, it takes years to do all that and in the meantime you're wasting energy on not staying at the forefront of a hot tech trend.

link