BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260522T150116Z
LOCATION:D166
DTSTART;TZID=America/Chicago:20181112T143000
DTEND;TZID=America/Chicago:20181112T150000
UID:submissions.supercomputing.org_SC18_sess173_ws_espm109@linklings.com
SUMMARY:A Unified Runtime for PGAS and Event-Driven Programming
DESCRIPTION:Sri Raj Paul (Rice University), Kun Chen (Georgia Institute of
  Technology), Akihiro Hayashi and Max Grossman (Rice University), and Vive
 k Sarkar (Georgia Institute of Technology)\n\nA well-recognized characteri
 stic of extreme scale systems is that their computation bandwidths far exc
 eed their communication bandwidths.  PGAS runtimes have proven to be effec
 tive in enabling efficient use of communication bandwidth, due to their ef
 ficient support for short nonblocking one-sided messages.  However, they w
 ere not designed for exploiting the massive levels of intra-node paralleli
 sm found in extreme scale systems.\n\nIn this paper, we explore the premis
 e that event-driven intra-node runtimes could be promising candidates for 
 integration with PGAS runtimes, due to their ability to overlap computatio
 n with long-latency operations.  For this, we use OpenSHMEM as an exemplar
  of PGAS runtimes, and Node.js as an exemplar of event-driven runtimes.  W
 hile Node.js may seem an unusual choice for high-performance computing, it
 s prominent role as an event-driven runtime for server-side Javascript is 
 a good match for what we need for optimizing the use of communication band
 width.  Availability of excess computation bandwidth in extreme scale syst
 ems can help mitigate the local computation overheads of Javascript.  Furt
 her, since the Node.js environment is single-threaded, we get an automatic
  guarantee that no data races will occur on Javascript objects, a guarante
 e that we cannot get with C++.\n\nOur integration of OpenSHMEM and Node.js
  makes it possible to expose nonblocking PGAS operations as JavaScript con
 structs that can seamlessly be used with the JavaScript asynchronous mecha
 nisms, such as those based on promises.  We believe that the exploration a
 nd preliminary results in this paper offer a new direction for future rese
 arch on building high productivity runtimes for extreme scale systems.\n\n
 Tag: Accelerators, Exascale, Parallel Programming Languages, Libraries, an
 d Models\n\nRegistration Category: Workshop Reg Pass\n\nSession Chairs: Dh
 abaleswar K. (DK) Panda (The Ohio State University), Karl Schulz (Advanced
  Micro Devices (AMD) Inc), and Hari Subramoni (The Ohio State University)\
 n\n
END:VEVENT
END:VCALENDAR