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:20181221T160906Z
LOCATION:D163
DTSTART;TZID=America/Chicago:20181112T090000
DTEND;TZID=America/Chicago:20181112T173000
UID:submissions.supercomputing.org_SC18_sess142@linklings.com
SUMMARY:PDSW-DISCS: Joint International Workshop on Parallel Data Storage 
 and Data Intensive Scalable Computing Systems
DESCRIPTION:Workshop\nI/O, Storage, Workshop Reg Pass\n\nPufferbench: Eval
 uating and Optimizing Malleability of Distributed Storage\n\nCheriere, Dor
 ier, Antoniu\n\nMalleability is the property of an application to be dynam
 ically rescaled at run time. It requires the possibility to dynamically ad
 d or remove resources to the infrastructure without interruption. Yet, man
 y Big Data applications cannot benefit from their inherent malleability, s
 ince their colocated...\n\n---------------------\nEvaluation of HPC Applic
 ation I/O on Object Storage Systems\n\nLiu, Koziol, Butler, Fortner, Chaar
 awi...\n\nPOSIX-based parallel file systems provide strong consistency sem
 antics, which many modern HPC applications do not need and do not want.  O
 bject store technologies avoid POSIX consistency and are designed to be ex
 tremely scalable, for use in cloud computing and similar commercial enviro
 nments.  In th...\n\n---------------------\nWelcome and Introduction\n\n\n
 \n---------------------\nWIP Session 2\n\n\n\n---------------------\nWIP S
 ession 3\n\n\n\n---------------------\nUnderstanding SSD Reliability in La
 rge-Scale Cloud Systems\n\nXu, Zheng, Qin, Xu, Wu\n\nModern datacenters in
 creasingly use flash-based solid state drives (SSDs) for high performance 
 and low energy cost. However, SSDs introduce more complex failure modes co
 mpared to traditional hard disks. While great efforts have been made to un
 derstand the reliability of SSDs itself, it remains uncle...\n\n----------
 -----------\nIntroduction - PDSW-DISCS: Joint International Workshop on Pa
 rallel Data Storage and Data Intensive Scalable Computing Systems\n\nMohro
 r, McIntosh, Raja, Maltzahn, Jimenez...\n\nWe are pleased to announce that
  the third Joint International Workshop on Parallel Data Storage and Data 
 Intensive Scalable Computing Systems (PDSW-DISCS’18) will be hosted at SC1
 8: The International Conference for High Performance Computing, Networking
 , Storage, and Analysis.  The objective of this...\n\n--------------------
 -\nWorkshop Lunch (on your own)\n\n\n\n---------------------\nWorkshop Mor
 ning Break\n\n\n\n---------------------\nKeynote Address\n\nSukumar\n\n---
 ------------------\nToward Understanding I/O Behavior in HPC Workflows\n\n
 Luettgau, Snyder, Carns, Wozniak, Kunkel...\n\nScientific discovery increa
 singly depends on complex workflows consisting of multiple phases and some
 times millions of parallelizable tasks or pipelines. These workflows acces
 s storage resources for a variety of purposes, including preprocessing, si
 mulation output, and postprocessing steps. Unfortun...\n\n----------------
 -----\nMethodology for the Rapid Development of Scalable HPC Data Services
 \n\nDorier, Carns, Harms, Latham, Ross...\n\nGrowing evidence in the scien
 tific computing community indicates that parallel file systems are not suf
 ficient for all HPC storage workloads.  This realization has motivated ext
 ensive research in new storage system designs. The question of which desig
 n we should turn to implies that there could be a...\n\n------------------
 ---\nWIP Session 1\n\n\n\n---------------------\nWorkshop Afternoon Break\
 n\n\n\n---------------------\nCharacterizing Deep-Learning I/O Workloads i
 n TensorFlow\n\nChien, Markidis, Sishtla, Santos, Herman...\n\nThe perform
 ance of Deep-Learning (DL) computing frameworks rely on the performance of
  data ingestion and checkpointing. In fact, during the training, a conside
 rable high number of relatively small files are first loaded and pre-proce
 ssed on CPUs and then moved to accelerator for computation. In addi...\n\n
 ---------------------\nIntegration of Burst Buffer in High-Level Parallel 
 I/O Library for Exascale Computing Era\n\nHou, Al-Bahrani, Rangel, Agrawal
 , Latham...\n\nWhile the computing power of supercomputers continues to im
 prove at an astonishing rate, companion I/O systems are struggling to keep
  up in performance. To mitigate the performance gap, several supercomputin
 g systems have been configured to incorporate burst buffers into their I/O
  stack; the exact r...\n\n---------------------\nUsing a Robust Metadata M
 anagement System to Accelerate Scientific Discovery at Extreme Scales\n\nL
 awson, Lofstead\n\nOur previous work, which can be referred to as EMPRESS 
 1.0, showed that rich metadata management provides a relatively low-overhe
 ad approach to facilitating insight from scale-up scientific applications.
  However, this system did not provide the functionality needed for a viabl
 e production system or ...\n
END:VEVENT
END:VCALENDAR