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:tcg40's list
SUMMARY:CASCADE Poster Social (+ Compiler Tech Talk) - Sea
 n Silva\, Encharge AI
DTSTART;TZID=Europe/London:20250311T150000
DTEND;TZID=Europe/London:20250311T180000
UID:TALK228928AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/228928
DESCRIPTION:*"CASCADE":https://www.cl.cam.ac.uk/research/casca
 de/people/ Poster Social*\n\n\nPlease join us for 
 a look at some of the cutting-edge research in com
 puter architecture coming out of the Department of
  Computer Science and Technology with a poster ses
 sion and light refreshments. We'll have posters fr
 om students and postdocs working in the broad area
  of computer architecture\, where you can browse a
 t your leisure\, speak directly to the authors and
  network with other computer architects whilst enj
 oying an early evening drink and light snacks in t
 he William Gates Building.\n\n**Schedule**\n\n15:0
 0 - 16:00 | Compiler Tech Talks   \n\n16:00 - 18:0
 0 | Cascade Poster Social\n\n**Registration**\nhtt
 ps://www.eventbrite.co.uk/e/cascade-poster-session
 -and-social-event-tickets-1249170325069\n\nBefore 
 the Cascade Poster Social\, we are hosting a Compi
 ler Tech Talk with Sean Silva.\n\n**A High-Velocit
 y Architecture for MLIR AI Compilers**\n\n\nIn thi
 s talk we present an MLIR-based compiler architect
 ure that delivers predictable performance for AI w
 orkloads. We derive its design from intuitive ment
 al models\, and along the way subsume seemingly co
 mpeting goals of performance\, portability\, and g
 enerality into one concept: velocity (of developin
 g the compiler). Hence we call this a high-velocit
 y compiler architecture. We present this compiler 
 architecture in detail\, compare it with the most 
 widespread competing architectural ideas\, and con
 nect it with prior art.\n \n**Sean Silva**\n\n... 
 is a Principal Engineer at Encharge AI\, where amo
 ng other responsibilities he leads the architectur
 e and implementation of the compiler. He previousl
 y worked on state-of-the-art AI software and hardw
 are for 7 years at Google and Waymo\, including IR
 EE\, XLA\, Google Edge TPU\, and Pixel Visual Core
 . His humble start in the LLVM community was refac
 toring TableGen in 2012 which has grown into a dec
 ade-long adventure through LLVM\, Clang\, LLD\, an
 d MLIR.\n\nhttps://www.linkedin.com/in/sean-silva-
 144b611b5/\n\nSlides: https://gist.github.com/ssil
 va-encharge/022659297610a763792c45d56dad49e3\n
LOCATION:Computer Laboratory\, William Gates Building\, LT2
CONTACT:Tobias Grosser
END:VEVENT
END:VCALENDAR
