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:20250822T115804Z
LOCATION:Campussaal - Plenary Room
DTSTART;TZID=Europe/Stockholm:20250616T102000
DTEND;TZID=Europe/Stockholm:20250616T105000
UID:submissions.pasc-conference.org_PASC25_sess149_pos104@linklings.com
SUMMARY:P26 - Improving Productivity of Threaded Scientific Applications w
 ith Quo Vadis
DESCRIPTION:Edgar A. Leon (Lawrence Livermore National Laboratory); Samuel
  K. Gutierrez (Los Alamos National Laboratory); and Guillaume Mercier (Bor
 deaux-INP; Inria, CNRS, LaBRI UMR 5800)\n\nScientific discovery is increas
 ingly enabled by heterogeneous hardware that includes multiple processor t
 ypes, deep memory hierarchies, and heterogeneous memories. To effectively 
 utilize this hardware, computational scientists must compose their applica
 tions using a combination of programming models, middleware, and runtime s
 ystems. Since these systems are often designed in isolation from each othe
 r, their concurrent execution results in resource contention and interfere
 nce, which limits application performance and scalability. Consequently, r
 eal-world applications face numerous challenges on heterogeneous machines.
  This poster presents the thread interface of Quo Vadis, a runtime system 
 that helps hybrid applications make efficient use of\nheterogeneous hardwa
 re, eases programmability in the presence of multiple programming abstract
 ions, and enables portability across systems. Applications using OpenMP or
  POSIX threads can now benefit from Quo Vadis' high-level abstractions to 
 map and remap full physics\npackages to the machine dynamically with a han
 dful of functions. Furthermore, the thread interface has similar semantics
  to the process interface, allowing scientists to leverage a single-semant
 ics model for partitioning and assignment of resources to workers, whether
  they are processes or threads. With both process and thread interfaces, Q
 uo Vadis broadens its applicability to improve the productivity of computa
 tional scientists across programming abstractions and heterogeneous hardwa
 re.\n\nSession Chair: Chris Cantwell (Imperial College London)\n\n
END:VEVENT
END:VCALENDAR
