BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20191206T142455Z
LOCATION:Booth N-230
DTSTART;TZID=Europe/Stockholm:20180626T151500
DTEND;TZID=Europe/Stockholm:20180626T154500
UID:isc_hpc_ISC High Performance 2018_sess144_proj113@linklings.com
SUMMARY:(PP04) The High-Q Club
DESCRIPTION:Project Poster\n\n(PP04) The High-Q Club\n\nBrömmel, Frings, W
 ylie\n\nThis poster summarises knowledge gained from running and tuning hi
 ghly-scalable applications on JUQUEEN, the IBM Blue Gene/Q at Jülich Super
 computing Centre.  The ability to execute successfully on all 458,752 core
 s with up to 1.8 million processes or threads could qualify codes for the 
 High-Q Club, which serves as a showcase for diverse codes effectively defi
 ning a collection of the highest scaling codes on JUQUEEN.  The intention 
 was to encourage other application developers to invest in tuning and scal
 ing their codes while identifying the necessary aspects for that goal.  At
  the close of this era with the decommissioning of JUQUEEN in spring 2018,
  it is timely to compare the characteristics of the 32 High-Q Club member 
 codes, considering their strong and/or weak scaling, exploitation of hardw
 are threading, and whether/how intra-node multi-threading is employed comb
 ined with message-passing.    Standard programming languages and MPI combi
 ned with\nmulti-threading was found to be sufficient, and provided a strai
 ghtforward migration path for application developers supported by scalable
  tools and optimised libraries which has also delivered performance and sc
 alability benefits on diverse HPC computing systems.  Obstacles for scalin
 g such as inefficient use of limited compute node memory and file I/O are 
 identified as key governing factors.   Details are discussed in a review a
 rticle [DOI: 10.14529/js180104].  Overall, the analysis provides guidance 
 as to how applications may (need to) be designed or adapted to exploit exp
 ected exa-scale computer systems, although there are likely to be signific
 ant additional aspects such as processor heterogeneity and variability.  W
 e are considering how the High-Q Club should evolve for assessing extremel
 y scalable applications on successor leadership computer systems, and invi
 te suggestions from the HPC community.\n\nPasses: Conference Pass, Exhibit
 ion Pass, Extreme-Scale Computing, Parallel Applications, Performance Anal
 ysis and Optimization\n\nTag: Conference Pass, Exhibition Pass, Extreme-Sc
 ale Computing, Parallel Applications, Performance Analysis and Optimizatio
 n
URL:https://2018.isc-program.com/presentation/?id=proj113&sess=sess144
END:VEVENT
END:VCALENDAR