BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Naiad: a system for incremental\, iterative and interactive parall
 el computation - Frank McSherry (MSR Silicon Valley)
DTSTART:20121107T161500Z
DTEND:20121107T170000Z
UID:TALK41045@talks.cam.ac.uk
CONTACT:Eiko Yoneki
DESCRIPTION:We are developing a new system for large-scale data analysis -
 - called "Naiad" -- which has the goal of supporting complex iterative que
 ries over dynamic inputs at interactive timescales. Like many existing sys
 tems\, Naiad supports high-level declarative queries\, data-parallel execu
 tion\, and transparent distribution. Unlike these systems\, Naiad can effi
 ciently execute queries with multiple (possibly nested) iterative loops\, 
 while simultaneously supporting low-latency incremental changes to the que
 ry inputs. As a highlight of its characteristics\, Naiad can not only effi
 ciently compute the strongly connected component structure of a 24 hour sl
 iding window of the Twitter @mention graph (using a doubly nested fixed-po
 int computation)\, but also maintains the computation with sub-second late
 ncies in the face of Twitter's full volume of continuously arriving tweets
 .\n\nI will describe the computational model underlying Naiad\, a generali
 zation of traditional incremental dataflow to partially ordered logical ti
 mes\, and work through some of the (very friendly\, picture oriented) math
 ematical details. I will also highlight several new distributed systems ch
 allenges faced in order to fully realize the multiple orders-of-magnitude 
 performance improvements Naiad presents.\n\nThis is joint work with Derek 
 Murray\, Rebecca Isaacs\, Michael Isard\, and Martìn Abadi.\n\nBio:\nFran
 k McSherry is a Senior Researcher at Microsoft's Silicon Valley Lab\, wher
 e he focuses on issues related to large-scale data analysis. He has previo
 usly worked on machine learning and privacy issues\, and is currently hard
  at work on large-scale low-latency computational infrastructure.\n\n
LOCATION:FW26\, Computer Laboratory\, William Gates Builiding
END:VEVENT
END:VCALENDAR
