BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Scaling AI Systems with Optical I/O - Manya Ghobadi\, MIT
DTSTART:20200910T140000Z
DTEND:20200910T150000Z
UID:TALK149599@talks.cam.ac.uk
CONTACT:Srinivasan Keshav
DESCRIPTION:The emergence of optical I/O chiplets enables compute/memory c
 hips to communicate with several Tbps bandwidth. Many technology trends po
 int to the arrival of optical I/O chiplets as a key industry inflection po
 int to realize fully disaggregated systems. In this talk\, I will focus on
  the potential of optical I/O-enabled accelerators for building high bandw
 idth interconnects tailored for distributed machine learning training. Our
  goal is to scale the state-of-the-art ML training platforms\, such as NVI
 DIA's DGX\, from a few tightly connected GPUs in one package to hundreds o
 f GPUs while maintaining Tbps communication bandwidth across the chips. Ou
 r design enables accelerating the training time of popular ML models using
  a device placement algorithm that partitions the training job with data\,
  model\, and pipeline parallelism across nodes\, while ensuring a sparse a
 nd local communication pattern that can be supported efficiently on the in
 terconnect.\n\n\nBio: Manya Ghobadi is an assistant professor at the EECS 
 department at MIT. Before MIT\, she was a researcher at Microsoft Research
  and a software engineer at Google Platforms. Manya is a computer systems 
 researcher with a networking focus and has worked on a broad set of topics
 \, including data center networking\, optical networks\, transport protoco
 ls\, and network measurement. Her work has won the best dataset award and 
 best paper award at the ACM Internet Measurement Conference (IMC) as well 
 as Google research excellent paper award.\n
LOCATION:https://meet.google.com/ehj-dwaz-rea
END:VEVENT
END:VCALENDAR