Skip to main content
Publication

167-pflops Deep Learning for Electron Microscopy: From Learning Physics to Atomic Manipulation

Publication Type
Conference Paper
Journal Name
Proceedings of the International Conference for High Performance Computing, Networking, Storage, and Analysis
Publication Date
Page Numbers
1 to 11
Publisher Location
New York, New York, United States of America
Conference Name
International Conference for High Performance Computing, Networking, Storage, and Analysis (SC 18)
Conference Location
Dallas, Texas, United States of America
Conference Sponsor
ACM and IEEE
Conference Date
-

An artificial intelligence system called MENNDL, which used 25,200 NVIDIA Volta GPUs on Oak Ridge National Laboratory's Summit machine, automatically designed an optimal deep learning network in order to extract structural information from raw atomic-resolution microscopy data. In a few hours, MENNDL creates and evaluates millions of networks using a scalable, parallel, asynchronous genetic algorithm augmented with a support vector machine to automatically find a superior deep learning network topology and hyper-parameter set than a human expert can find in months. For the application of electron microscopy, the system furthers the goal of improving our understanding of the electron-beam-matter interactions and real-time image-based feedback, which enables a huge step beyond human capacity towards nanofabricating materials automatically. MENNDL has been scaled to the 4,200 available nodes of Summit achieving a measured 152.5 PFlops, with an estimated sustained performance of 167 PFlops when the entire machine is available.