BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Scale-Out Processors - Boris Grot\, University of Edinburgh
DTSTART:20131008T110000Z
DTEND:20131008T120000Z
UID:TALK48112@talks.cam.ac.uk
CONTACT:Robert Mullins
DESCRIPTION:Big data is revolutionizing the way we live\, work\, and socia
 lize. At the same time\, big data is taxing our compute infrastructure in 
 unprecedented ways. In many domains\, data expansion rates are dwarfing th
 e pace of technology improvement as measured by Moore’s law\, challengin
 g our ability to effectively store and process the data. Moreover\, with t
 he hardware industry hitting fundamental limits on its ability to lower op
 erating voltages\, energy requirements in big-data applications are skyroc
 keting. Sustaining the pressure of big data\,\nand delivering on its promi
 ses\, requires significant improvements in the efficiency of our compute i
 nfrastructure.\n\nIn this talk\, I will focus on data-intensive online app
 lications\, such as web search and social connectivity. I will explain how
  the mismatch\nbetween application demands and existing processor architec
 tures leads to significant inefficiencies at the datacenter level. I will 
 then describe Scale-Out Processors\, a processor design methodology and\nm
 icroarchitectural support for data-intensive online processing. By tuning 
 the processor organization to the needs of the application domain\, Scale-
 Out Processors improve datacenter performance by up to\n7x within a fixed 
 power budget versus state-of-the-art server processors.\n\nBio\n\nBoris Gr
 ot is a Lecturer in the School of Informatics at the University of Edinbur
 gh. His research seeks to address efficiency bottlenecks and capability sh
 ortcomings of processing platforms for big data. Grot received his PhD in 
 Computer Science from The University of Texas at Austin in 2011.\n\nhttp:/
 /parsa.epfl.ch/~grot/pubs.html
LOCATION:FW11\, Computer Laboratory
END:VEVENT
END:VCALENDAR