BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260522T150116Z
LOCATION:A2 Ballroom
DTSTART;TZID=America/Chicago:20181114T160000
DTEND;TZID=America/Chicago:20181114T163000
UID:submissions.supercomputing.org_SC18_sess466_gb103@linklings.com
SUMMARY:167-PFlops Deep Learning for Electron Microscopy: From Learning Ph
 ysics to Atomic Manipulation
DESCRIPTION:Robert M. Patton, J. Travis Johnston, Steven R. Young, Catheri
 ne D. Schuman, Don D. March, Thomas E. Potok, Derek C. Rose, Seung-Hwan Li
 m, Thomas P. Karnowski, Maxim A. Ziatdinov, and Sergei V. Kalinin (Oak Rid
 ge National Laboratory)\n\nAn artificial intelligence system called MENNDL
 , which used 25,200 Nvidia Volta GPUs on Oak Ridge National Laboratory’s S
 ummit machine, automatically designed an optimal deep learning network in 
 order to extract structural information from raw atomic-resolution microsc
 opy data. In a few hours, MENNDL creates and evaluates millions of network
 s using a scalable, parallel, asynchronous genetic algorithm augmented wit
 h a support vector machine to automatically find a superior deep learning 
 network topology and hyper-parameter set than a human expert can find in m
 onths. For the application of electron microscopy, the system furthers the
  goal of improving our understanding of the electron-beam-matter interacti
 ons and real-time image-based feedback, which enables a huge step beyond h
 uman capacity toward nanofabricating materials automatically. MENNDL has b
 een 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.\n\n
END:VEVENT
END:VCALENDAR