In this video, Tor Skeie from Simula Labs presents: Dragonfly versus Fat-tree — the future of cluster topologies. Recorded at the HPC Advisory Council European Workshop at ISC’12 in Hamburg.
A cumulative effort over several years to scale the training of deep-learning neural networks has resulted in the first demonstration of petascale deep-learning training performance, and further to deliver this performance when solving real science problems. The result reflects the combined efforts of NERSC (National Energy Research Scientific Computing Center), Stanford and Intel to solve real world use cases rather than simply report on performance benchmarks. [Read More...]
Using Remote Direct Memory Access based analytics and fast, scalable,external disk systems with massively parallel access to data, SAS analytics driven organizations can deliver timely and accurate execution for data intensive workflows such as risk management, while incorporating larger datasets than using traditional NAS.
Copyright © 2017
I recall reading this paper a couple of years ago. The interesting point at the end was (paraphrased) ‘hey: no hardware exists that supports our idea’. This hasn’t changed yet, has it?
I recall reading about this a couple of years back after someone had a talk with me about it at an SC. (cool Tshirt from the hpc lab in NO too) They problem was (paraphrased) ‘no hardware exists that supports this idea’. Has this changed?