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:20181221T160731Z
LOCATION:Exhibit Hall B
DTSTART;TZID=America/Chicago:20181115T111500
DTEND;TZID=America/Chicago:20181115T120000
UID:submissions.supercomputing.org_SC18_sess230_inv111@linklings.com
SUMMARY:Superscaling Performance through Energy-Efficient Supercomputing
DESCRIPTION:Invited Talk\nTech Program Reg Pass, Exhibits Reg Pass\n\nSupe
 rscaling Performance through Energy-Efficient Supercomputing\n\nRaghavan\n
 \nIn June 2008, Roadrunner sped past the 1 petaflop/s milestone, clocking 
 in at 1.026 petaflops/s. Ten short years later in June 2018, Summit has no
 w taken over the as the world’s fastest supercomputer spot by achieving 12
 2.3 petaflops/s. What is remarkable is that the over 100x scaling up of pe
 rformance was achieved with less than 4x scaling of hardware power. In thi
 s talk, I will highlight the key advances in energy-efficiency that made p
 ossible this superscaling of performance with a particular focus on the in
 fluence of linear algebra and graph algorithms. I will also speculate on t
 rajectories for their continued development to meet the demands of data ce
 ntric computing as we face slow-downs in reducing the power consumption of
  transistors.
URL:https://sc18.supercomputing.org/presentation/?id=inv111&sess=sess230
END:VEVENT
END:VCALENDAR