TY - GEN
T1 - 167-PFlops deep learning for electron microscopy
T2 - 2018 International Conference for High Performance Computing, Networking, Storage, and Analysis, SC 2018
AU - Patton, Robert M.
AU - Travis Johnston, J.
AU - Young, Steven R.
AU - Schuman, Catherine D.
AU - March, Don D.
AU - Potok, Thomas E.
AU - Rose, Derek C.
AU - Lim, Seung Hwan
AU - Karnowski, Thomas P.
AU - Ziatdinov, Maxim A.
AU - Kalinin, Sergei V.
N1 - Publisher Copyright:
© 2018 IEEE.
PY - 2018/7/2
Y1 - 2018/7/2
N2 - 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.
AB - 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.
KW - Evolutionary computation
KW - High performance computing
KW - Machine learning
UR - http://www.scopus.com/inward/record.url?scp=85064129765&partnerID=8YFLogxK
U2 - 10.1109/SC.2018.00053
DO - 10.1109/SC.2018.00053
M3 - Conference contribution
AN - SCOPUS:85064129765
T3 - Proceedings - International Conference for High Performance Computing, Networking, Storage, and Analysis, SC 2018
SP - 638
EP - 648
BT - Proceedings - International Conference for High Performance Computing, Networking, Storage, and Analysis, SC 2018
PB - Institute of Electrical and Electronics Engineers Inc.
Y2 - 11 November 2018 through 16 November 2018
ER -