BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:Computer Laboratory Systems Research Group Seminar
SUMMARY:Building modern dataflow systems - Frank McSherry 
 (ETH Zürich)
DTSTART;TZID=Europe/London:20180712T110000
DTEND;TZID=Europe/London:20180712T120000
UID:TALK108241AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/108241
DESCRIPTION:\nAbstract: I'll talk through the design and imple
 mentation of "timely dataflow in Rust"\, an open-s
 ource project that extends and enriches the "timel
 y dataflow" computational model first presented by
  the Naiad system\, and the differential dataflow 
 framework built on top of it. The project's goal i
 s to provide an near-zero overhead framework for d
 ata-parallel dataflow computation\, and to this en
 d it simplifies and unifies several of Naiad's con
 cepts through lossless abstractions that largely c
 ompile away. Our experience has been that timely d
 ataflow programs give best-in-class performance\, 
 while still providing the experience of a medium-t
 o-high level programming language. To support this
 \, I'll walk through the example of differential d
 ataflow\, an incremental re-computation framework 
 which seems to out-perform the current crop of spe
 cialized data processing systems\, in part due to 
 its ability to provide general computation abstrac
 tions that compile down to sequential scans over c
 arefully managed resources.\n\n	https://github.com
 /frankmcsherry/timely-dataflow\n	https://github.co
 m/frankmcsherry/differential-dataflow\n\nThese pro
 jects reflect joint work with a great many people\
 , including what was once the Naiad team at MSR-SV
 \, the Systems Group at ETH Zürich\, and many othe
 r collaborators.\n\nBio: Frank McSherry received h
 is PhD from the University of Washington\, working
  with Anna Karlin on spectral analysis of data. He
  then spent twelve years at Microsoft Research's S
 ilicon Valley research center\, working on topics 
 ranging from differential privacy to data-parallel
  computation. He currently works at ETH Zürich's S
 ystems Group on scalable stream processing and rel
 ated topics.
LOCATION:SW00\, Computer Laboratory\, William Gates Buildin
 g
CONTACT:Marco Caballero
END:VEVENT
END:VCALENDAR