BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/New_York
X-LIC-LOCATION:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20210402T160559Z
LOCATION:Track 8
DTSTART;TZID=America/New_York:20201113T143500
DTEND;TZID=America/New_York:20201113T145500
UID:submissions.supercomputing.org_SC20_sess227_ws_pyhpc101@linklings.com
SUMMARY:Enabling System Wide Shared Memory for Performance Improvement in 
 PyCOMPSs Applications
DESCRIPTION:Workshop\n\nEnabling System Wide Shared Memory for Performance
  Improvement in PyCOMPSs Applications\n\nFoyer, Conejero, Ejarque, Badia, 
 Tate...\n\nPython has been gaining some traction for years in the world of
  scientific applications. The high-level abstraction it provides, however,
  may not allow the developer to use the machines to their peak performance
 . To address this, multiple strategies, sometimes complementary, have been
  developed to enrich the software ecosystem either by relying on additiona
 l libraries dedicated to efficient computation (e.g., NumPy) or by providi
 ng a framework to better use HPC scale infrastructures (e.g., PyCOMPSs).\n
 \nIn this paper, we present a Python extension based on SharedArray that e
 nables the support of system-provided shared memory and its integration in
 to the PyCOMPSs programming model as an example of integration to a comple
 x Python environment. We also evaluate the impact such a tool may have on 
 performance in two types of distributed execution-flows, one for linear al
 gebra with a blocked matrix multiplication application and the other in th
 e context of data-clustering with a k-means application. We show that with
  very little modification of the original decorator (three lines of code t
 o be modified) of the task-based application the gain in performance can r
 ise above 40% for tasks relying heavily on data reuse on a distributed env
 ironment, especially when loading the data is prominent in the execution t
 ime.\n\nRegistration Category: Workshop Reg Pass
END:VEVENT
END:VCALENDAR