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:20210916T132527Z
LOCATION:Mère Royaume
DTSTART;TZID=Europe/Stockholm:20210706T140000
DTEND;TZID=Europe/Stockholm:20210706T160000
UID:submissions.pasc-conference.org_PASC21_sess137@linklings.com
SUMMARY:Data Movement Orchestration on HPC Systems
DESCRIPTION:Minisymposium\n\nThe next HPC systems are expected to feature 
 millions of cores leading to more parallelism for large-scale simulations 
 and workflows in varied scientific domains. However, there is a price to p
 ay: the gap between computing power and data movement performance will kee
 p increasing more and more and will inevitably intensify the bottleneck ca
 used by data movements. Thus, it becomes necessary to consider data moveme
 nt within those modern architectures at all the stages of the data lifetim
 e and to come up with software components for orchestrating this movement.
  In particular, we can identify several key aspects among which: data sema
 ntics for expressing applications needs and memory abstraction for portabi
 lity and extendibility given the multiplicity of memory and storage tiers.
  During this Minisymposium, we will focus on those two facets and present 
 very promising ongoing projects coming from academia and industry proposin
 g to address the data orchestration challenge and facilitate efficient exp
 loitation of future exascale systems. An open discussion with the speakers
  will close the session.\n\nApplication Coordination in Data-Driven Workfl
 ows\n\nHaus, Haine\n\nMoving data through memory has been recognized as a 
 central performance bottleneck in the exascale era. This is of particular 
 relevance when coupling multiple applications in workflows, where quite of
 ten I/O to disc in serialized form, or specialized domain-specific applica
 tion couplers are two extr...\n\n---------------------\nFrom Monitoring to
  Process Placement a Complete Workflow for Optimizing Data Management at S
 ystem Scale\n\nJeannot\n\nData management and especially communication opt
 imization is one of the key techniques to achieve performance and scalabil
 ity of high-performance computing applications. One of the interesting tec
 hniques for minimizing communication cost and data exchange between proces
 ses is to perform a careful pr...\n\n---------------------\nAutomatic and 
 Transparent Scientific Data Management with Object Abstractions\n\nByna, K
 oziol, Tang\n\nThree trends are dominating scientific data management: (1)
  large volumes of data (2) massive concurrency of supercomputers, and (3) 
 deepening hierarchy of heterogeneous memory and storage devices. For decad
 es, storage systems have performed I/O with file-oriented approaches based
  on the POSIX and M...\n\n---------------------\nPanel Discussion on Data 
 Movement Orchestration on HPC Systems\n\nTessier\n\nThe goal of the three 
 talks of this minisymposium was to give an overview of the issues surround
 ing the orchestration of data movement on large scale infrastructures. The
  new machines that will be deployed this year and that will likely reach t
 he Exascale, come with new issues about how to manage da...\n\n\nDomain: C
 S and Math
END:VEVENT
END:VCALENDAR
