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:20241120T082410Z
LOCATION:HG F 30 Audi Max
DTSTART;TZID=Europe/Stockholm:20240603T190800
DTEND;TZID=Europe/Stockholm:20240603T191000
UID:submissions.pasc-conference.org_PASC24_sess156_pos105@linklings.com
SUMMARY:P19 - Fast and Scalable Algorithms for Selected Matrix Inversions
DESCRIPTION:Poster\n\nLisa Gaedke-Merzhäuser (Università della Svizzera it
 aliana); Vincent Maillou, Alexandros N. Ziogas, and Mathieu Luisier (ETH Z
 urich); and Olaf Schenk (Università della Svizzera italiana)\n\nThe invers
 ion of sparse linear systems gives rise to dense matrices. Their computati
 on poses not only a computational but also a memory bottleneck. Numerous a
 pplications from various fields require, however, only particular, i.e. se
 lected entries of the complete inverse. Applications range from areas like
  statistical learning where the computation of marginal variances requires
  the selected inversion of the associated sparse precision matrices, to na
 no-electronics in device physics, where quantum transport simulations nece
 ssitate selected matrix inversions to model electron flows. The most commo
 n elements of interest in the inverse are the (block) diagonal elements or
  entries that correspond to non-zero elements in the original sparse matri
 x. We present a selected inversion algorithm for matrices with block tridi
 agonal arrowhead sparsity patterns which recovers all block diagonal entri
 es of the full inverse. Our implementation relies on block-wise dense GPU 
 computations and scales efficiently across multiple GPUs.\n\nSession Chair
 : Erik W. Draeger (Lawrence Livermore National Laboratory)
END:VEVENT
END:VCALENDAR
