BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20230831T095742Z
LOCATION:Davos
DTSTART;TZID=Europe/Stockholm:20230628T100000
DTEND;TZID=Europe/Stockholm:20230628T103000
UID:submissions.pasc-conference.org_PASC23_sess124_pap124@linklings.com
SUMMARY:AK03 - FourCastNet: Accelerating Global High-Resolution Weather Fo
 recasting Using Adaptive Fourier Neural Operators
DESCRIPTION:Keynote, Paper\n\nThorsten Kurth (NVIDIA Inc.); Shashank Subra
 manian and Peter Harrington (Lawrence Berkeley National Laboratory); and J
 aideep Pathak, Morteza Mardani, David Hall, Andrea Miele, Karthik Kashinat
 h, and Anima Anandkumar (NVIDIA Inc.)\n\nExtreme weather amplified by clim
 ate change is causing increasingly devastating impacts across the globe. T
 he current use of physics-based numerical weather prediction (NWP) limits 
 accuracy due to high computational cost and strict time-to-solution limits
 . We report that a data-driven deep learning Earth system emulator, FourCa
 stNet, can predict global weather and generate medium-range forecasts five
  orders-of-magnitude faster than NWP while approaching state-of-the-art ac
 curacy. FourCastNet is optimized and scales efficiently on three supercomp
 uting systems: Selene, Perlmutter, and JUWELS Booster up to 3,808 NVIDIA A
 100 GPUs, attaining 140.8 Petaflops in mixed precision (11.9% of peak at t
 hat scale). The time-to-solution for training FourCastNet measured on JUWE
 LS Booster on 3,072GPUs is 67.4minutes, resulting in an 80,000 times faste
 r time-to-solution relative to state-of-the-art NWP, in inference. FourCas
 tNet produces accurate instantaneous weather predictions for a week in adv
 ance and enables enormous ensembles that could be used to improve predicti
 ons of rare weather extremes.\n\nSession Chair: Cristina Silvano (Politecn
 ico di Milano)
END:VEVENT
END:VCALENDAR