Article doesn’t seem to discuss the results... would be interesting to see how the NN model did against traditional models
A TensorFlow model running on 27,000 GPUs at Oak Ridge National Laboratory, detecting extreme weather patterns. The world's fastest supercomputer.
This is what a powerful machine learning software platform looks like 
https://www.wired.com/story/worlds-fastest-supercomputer-breaks-ai-record/ …https://twitter.com/martin_wicke/status/1091146494202146816 …
-
-
-
I'd like to know this too.
End of conversation
New conversation -
-
-
Are GPUs in general winning out over bespoke supercomputers now ? Like Cray etc
-
Many of the top supercomputers use GPUs (including ones built by Cray)
End of conversation
New conversation -
-
-
You had me at tensors!
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
That's a lot of GPU
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Instory pic says 37000 processor n then writes 27000 graphic processor. Some gap? No results? It is fastest? How much? No benchmark! Wired again.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Super and how many days did you trained the model for predicting the output.
-
Not sure about Summit, but its predecessor (Titan) had a pretty brutal job time limit. If you couldn't aggressively trade time for more compute, you were out of luck.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.