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:20230124T171523Z
LOCATION:C146
DTSTART;TZID=America/Chicago:20221113T094000
DTEND;TZID=America/Chicago:20221113T100500
UID:submissions.supercomputing.org_SC22_sess424_ws_scalah102@linklings.com
SUMMARY:GPU Optimization of Lattice Boltzmann Method with Local Ensemble T
ransform Kalman Filter
DESCRIPTION:Workshop\n\nGPU Optimization of Lattice Boltzmann Method with
Local Ensemble Transform Kalman Filter\n\nHasegawa, Imamura, Ina, Onodera,
Asahi...\n\nThe ensemble data assimilation of computational fluid dynamic
s simulations based on the lattice Boltzmann method (LBM) and the local en
semble transform Kalman filter (LETKF) is implemented and optimized on a G
PU supercomputer based on NVIDIA A100 GPUs. To connect the LBM and LETKF p
arts, data transpose communication is optimized by overlapping computation
, file I/O, and communication based on data dependency in each LETKF kerne
l. In two dimensional forced isotropic turbulence simulations with the ens
emble size of M=64 and the number of grid points of N_x=128^2, the optimiz
ed implementation achieved x3.80 speedup from the naive implementation, in
which the LETKF part is not parallelized. The main computing kernel of th
e local problem is the eigenvalue decomposition (EVD) of M x M real symmet
ric dense matrices, which is computed by a newly developed batched EVD in
EigenG. The batched EVD in EigenG outperforms that in cuSOLVER, and x65.3
speedup was achieved.\n\nSession Format: Recorded\n\nTag: Algorithms, Exas
cale Computing, Extreme Scale Computing, Heterogeneous Systems, Post-Moore
Computing, Quantum Computing\n\nRegistration Category: Workshop Reg Pass
END:VEVENT
END:VCALENDAR