BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VTIMEZONE
TZID:America/New_York
BEGIN:DAYLIGHT
DTSTART:20200308T030000
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZNAME:EDT
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20201101T010000
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZNAME:EST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20240527T022156Z
UID:11e2b705-2ee2-4554-a4f5-879c5d4c2354
DTSTART;TZID=America/New_York:20200810T140000
DTEND;TZID=America/New_York:20200810T150000
CREATED:20200803T151709
DESCRIPTION:Abstract: Algorithms have two costs: arithmetic and communicati
on\, i.e. moving data between levels of a memory hierarchy or processors o
ver a network. Communication costs (measured in time or energy per operati
on) already greatly exceed arithmetic costs\, and the gap is growing over
time following technological trends. Thus our goal is to design algorithms
that minimize communication. We present new algorithms that communicate a
symptotically less than their classical counterparts\, for a variety of li
near algebra and machine learning problems\, demonstrating large speedups
on a variety of architectures. Some of these algorithms attain provable lo
wer bounds on communication. We describe generalizations of these bounds\,
and optimal algorithms\, to arbitrary code that can be expressed as neste
d loops accessing arrays\, such as convolutional neural nets\, and to acco
unt for arrays having different precisions.\n\nBio: James Demmel is the Dr
. Richard Carl Dehmel Distinguished Professor of Computer Science and Math
ematics at the University of California at Berkeley\, and former Chair of
the EECS Dept. His research is in numerical linear algebra\, high performa
nce computing\, and communication avoiding algorithms. He is known for his
work on the widely used LAPACK and ScaLAPACK linear algebra libraries. H
e is a member of the National Academy of Sciences\, National Academy of En
gineering\, and American Academy of Arts and Sciences\; a Fellow of the AA
AS\, ACM\, AMS\, IEEE and SIAM\; and winner of the IPDPS Charles Babbage A
ward\, IEEE Computer Society Sidney Fernbach Award\, the ACM Paris Kanella
kis Award\, and numerous best paper prizes. \n\n
LAST-MODIFIED:20200803T170654
LOCATION:https://mit.zoom.us/meeting/register/tJUrdOqopj8uHdO4gUyVMnfglOFEq
Iye_Je0 (Registration required)
SUMMARY:Communication-avoiding algorithms for linear algebra\, machine lear
ning and beyond
URL:https://calendar.csail.mit.edu/events/235198
END:VEVENT
END:VCALENDAR