BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Denver
X-LIC-LOCATION:America/Denver
BEGIN:DAYLIGHT
TZOFFSETFROM:-0700
TZOFFSETTO:-0600
TZNAME:MDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0600
TZOFFSETTO:-0700
TZNAME:MST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20200129T163557Z
LOCATION:710
DTSTART;TZID=America/Denver:20191118T142000
DTEND;TZID=America/Denver:20191118T144000
UID:submissions.supercomputing.org_SC19_sess125_pec265@linklings.com
SUMMARY:Sub-Microsecond Optical Circuit Switched Networks for Data Centers
  and HPCs
DESCRIPTION:Workshop\n\nSub-Microsecond Optical Circuit Switched Networks 
 for Data Centers and HPCs\n\nZervas\n\nThere have been over two decades of
  research on optical switched technologies for Data Centers and HPCs. Howe
 ver there is still no commercial uptake and adoption by Cloud and HPC prov
 iders. This presentation will delve into why is that and elaborate on the 
 roadblocks, opportunities and choices to address the Moore’s Law slow down
  and replace electronic switching. It will explain the reasons why optical
  circuit switching rather than optical packet switching is the best way fo
 rward for Data Centers and HPCs and what are the main challenges. PULSE, a
  scalable network architecture that can reconfigure optical time-slotted c
 ircuits in nanoseconds will be introduced to support heterogeneous high pe
 rformance systems . An insight on technologies we are developing for sub-m
 icrosecond network scheduling and sub-nanosecond switching that aim to max
 imise throughout and minimise latency across thousands of end points (Serv
 ers, CPUs, accelerators, memory, etc.) will be provided. Crucially, I’ll p
 resent how the network can offer zero data loss, without in-network a) buf
 fering, b) active switching and routing, and c) network header addressing 
 and processing to minimize complexity, and crucially to consume very low p
 ower. Furthermore, the system also aims to inherently support 1-to-1, 1-to
 -N, N-to-N and N-to-1 connectivity in a synchronous manner without the nee
 d for data replication for multi/broad -casting, currently not possible. T
 his is key to support diverse workloads such as storage caching, large-sca
 le database lookups, training distributed DNNs, parallel computing that us
 e communication primitives such as allreduce, broadcast and reduce, gather
  and scatter, all-to-all among others.\n\nTag: Workshop Reg Pass, Architec
 tures, Datacenter, Emerging Technologies, Hardware, HPC, I/O, Networks, Ph
 otonics, Silicon Fabrication\n\nRegistration Category: Workshop Reg Pass, 
 Architectures, Datacenter, Emerging Technologies, Hardware, HPC, I/O, Netw
 orks, Photonics, Silicon Fabrication
URL:https://sc19.supercomputing.org/presentation/?id=pec265&sess=sess125
END:VEVENT
END:VCALENDAR