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:20260422T143047Z
LOCATION:B302
DTSTART;TZID=America/New_York:20241118T090000
DTEND;TZID=America/New_York:20241118T173000
UID:submissions.supercomputing.org_SC24_sess756@linklings.com
SUMMARY:The 19th Workshop on Workflows in Support of Large-Scale Science (
 WORKS24)
DESCRIPTION:Scientific workflows have underpinned some of the most signifi
 cant discoveries of the past several decades. Workflow management systems 
 (WMS) provide abstraction and automation that enable researchers to easily
  define sophisticated computational processes, and to then execute them ef
 ficiently on parallel and distributed computing systems. As workflows have
  been adopted by multiple scientific communities, they are becoming more c
 omplex and require more sophisticated workflow management capabilities. Th
 is workshop focuses on the many facets of scientific workflow composition,
  management, sustainability, and application to domain sciences in an incr
 easingly diverse landscape. The workshop covers a broad range of topics in
  the scientific workflow lifecycle that include: reproducible research wit
 h workflows; workflow execution in distributed and heterogeneous environme
 nts; application of AI/ML in workflow management; workflow provenance; ser
 verless workflows; exascale computing with workflows; stream-processing, i
 nteractive, adaptive and data-driven workflows; workflow scheduling and re
 source management; workflow fault-tolerance, debugging, performance analys
 is/modeling; big data and AI workflows, etc.\n\nAccelerating the Operation
  of Complex Workflows through Standard Data Interfaces\n\nIn this position
  paper we argue for standardizing how we share and process data in scienti
 fic workflows at the network-level to maximize step re-use and workflow po
 rtability across platforms and networks in pursuit of a foundational workf
 low stack. We look to evolve workflows from steps connected po...\n\n\nTay
 lor Paul and William Regli (University of Maryland)\n---------------------
 \nTowards a Cohesive Ecosystem of Workflows, Data, Artificial Intelligence
 , and Humans\n\nScientific discoveries increasingly depend on leveraging c
 omputation and data at scale among and across ecosystems. Scientific workf
 low tools provide a construct to manage the computation and data over dist
 ributed and large-scale infrastructure. In the panel discussion, we will d
 etail how the WORKS ...\n\n\nDrew Paine, Rajshree Deshmukh, Dan Gunter, Co
 dy O'Donnell, Sarah Poon, and Lavanya Ramakrishnan (Lawrence Berkeley Nati
 onal Laboratory (LBNL))\n---------------------\nAssessing and Advancing th
 e Potential of Quantum Computing: A NASA Case Study\n\nEleanor Rieffel (NA
 SA)\n---------------------\nWORKS24 — Lunch Break\n---------------------\n
 Shepherd: Seamless Integration of Service Workflows into Task-Based Workfl
 ows through Log Monitoring\n\nTraditional workflow managers focus on coord
 inating discrete tasks: actions that run to completion. However, emerging 
 workflows require persistent services that must be managed alongside tradi
 tional tasks. We introduce Shepherd, a local workflow manager that runs se
 rvices as a task, enabling them to...\n\n\nMd Saiful Islam and Douglas Tha
 in (University of Notre Dame)\n---------------------\nServerless Computing
  for Dynamic HPC Workflows\n\nContainers have become an important componen
 t for scientific workflows, enhancing reproducibility, portability, and is
 olation when coupled with workflow management systems. However, integratin
 g containers with these systems can be complex, potentially hindering wide
 r adoption. Serverless platforms o...\n\n\nVijay Thurimella (Georgia Insti
 tute of Technology, Hewlett Packard Enterprise (HPE))\n-------------------
 --\nPerformance Characterization and Provenance of Distributed Task-based 
 Workflows on HPC Platforms\n\nUnderstanding performance and provenance of 
 task-based workflows poses significant challenges, particularly in distrib
 uted configurations where resources are shared by multiple applications. T
 ask-based workflow management systems further complicate performance predi
 ctability because of their dynamic...\n\n\nAmal Gueroudji (Argonne Nationa
 l Laboratory (ANL)); Chase Phelps and Tanzima Z. Islam (Texas State Univer
 sity); Philip Carns, Shane Snyder, Matthieu Dorier, and Robert Ross (Argon
 ne National Laboratory (ANL)); and Line Pouchard (Sandia National Laborato
 ries)\n---------------------\nA software Ecosystem for Multi-Level Provena
 nce Management in Large-Scale Scientific Workflows for AI Applications\n\n
 Scientific workflows and provenance are two faces of the same medal. While
  the former addresses the coordinated execution of multiple tasks over a s
 et of computational resources, the latter relates to the historical record
  of data from its original sources. This paper highlights the importance o
 f tr...\n\n\nGabriele Padovani (University of Trento, Italy); Valentine An
 antharaj (Oak Ridge National Laboratory (ORNL)); Ludovica Sacco (Universit
 y of Trento, Italy); Takuya Kurihana (Oak Ridge National Laboratory (ORNL)
 ); Matteo Bunino, Kalliopi Tsolaki, and Maria Girone (European Organizatio
 n for Nuclear Research (CERN)); Fabrizio Antonio (Euro-Mediterranean Cente
 r on Climate Change Foundation); and Carolina Sopranzetti, Massimiliano Fr
 onza, and Sandro Fiore (University of Trento, Italy)\n--------------------
 -\nIntegrating Evolutionary Algorithms with Distributed Deep Learning for 
 Optimizing Hyperparameters on HPC Systems\n\nHigh performance computing (H
 PC) systems have become essential for solving complex scientific problems,
  particularly in the context of deep learning (DL). This extended abstract
  presents a novel system that uses a multiobjective evolutionary algorithm
  (EA) to optimize hyperparameters for a deep lear...\n\n\nMark Coletti, Re
 nan Santos Souza, Tyler Skluzacek, Frédéric Suter, and Rafael Ferreira da 
 Silva (Oak Ridge National Laboratory (ORNL))\n---------------------\nClosi
 ng\n\nSilvina Caino-Lores (French Institute for Research in Computer Scien
 ce and Automation (INRIA)) and Anirban Mandal (Renaissance Computing Insti
 tute (RENCI))\n---------------------\nPanel Discussion: Future of Scientif
 ic Workflows\n\nSilvina Caino-Lores (French Institute for Research in Comp
 uter Science and Automation (INRIA)) and Anirban Mandal (Renaissance Compu
 ting Institute (RENCI))\n---------------------\nWelcome\n\nSilvina Caino-L
 ores (French Institute for Research in Computer Science and Automation (IN
 RIA)) and Anirban Mandal (Renaissance Computing Institute (RENCI))\n------
 ---------------\nA Microservices Architecture Toolkit for Interconnected S
 cience Ecosystems\n\nMicroservices architecture is a promising approach fo
 r developing reusable scientific workflow capabilities for integrating div
 erse resources, such as experimental and observational instruments and adv
 anced computational and data management systems, across many distributed o
 rganizations and faciliti...\n\n\nMichael J. Brim, Lance Drane, Marshall M
 cDonnell, Christian Engelmann, and Addi Malviya Thakur (Oak Ridge National
  Laboratory (ORNL))\n---------------------\nTowards Generating Contracts f
 or Scientific Data Analysis Workflows\n\nTo increase the dependability and
  portability of scientific data analysis workflows (DAWs), recent work has
  proposed contract-driven design of DAWs, providing verifiable expectation
 s and obligations to ensure that tasks run in a proper environment and pro
 duce correct results.\nHowever, the specificat...\n\n\nAnh Duc Vu (Univers
 ity of Bern, Switzerland; Gesellschaft für Informatik) and Timo Kehrer (Un
 iversity of Bern, Switzerland)\n---------------------\nParsl+CWL: Towards 
 Combining the Python and CWL Ecosystems\n\nCommon Workflow Language (CWL) 
 is a widely adopted language for defining and sharing computational workfl
 ows. It is designed to be independent of the execution engine on which wor
 kflows are executed. Here, we describe our experiences integrating CWL wit
 h Parsl, a Python-based parallel programming li...\n\n\nNishchay Karle (Un
 iversity of Chicago, Globus); Ben Clifford (Hawaga); Yadu Babuji (Globus, 
 University of Chicago); Ryan Chard (Argonne National Laboratory (ANL)); Da
 niel S. Katz (University of Illinois Urbana-Champaign); and Kyle Chard (Gl
 obus, University of Chicago)\n---------------------\nLaminar 2.0: Serverle
 ss Stream Processing with Enhanced Code Search and Recommendations\n\nThis
  paper presents Laminar 2.0, an enhanced serverless framework for running 
 dispel4py streaming workflows. Building on Laminar, this version introduce
 s improved dependency management, client-server functionality, and advance
 d deep learning models for semantic search. Key innovations include a stru
 ...\n\n\nDaniel Rotchford and Samuel Evans (University of St Andrews, Scot
 land) and Rosa Filgueira (EPCC, The University of Edinburgh)\n------------
 ---------\nManaging Workflow Malleability in Urgent Computing for Earthqua
 ke Alerts\n\nWhen large earthquakes happen, first responders need fast and
  accurate information regarding their impact. UCIS4EQ is an urgent computi
 ng platform that estimates ground shaking based on high-performance parall
 el 3D simulations. In this work, we present the PyCOMPSs implementation of
  UCIS4EQ towards ...\n\n\nJorge Ejarque, Marisol Monterrubio-Velasco, and 
 Cedric Bhihe (Barcelona Supercomputing Center (BSC)); Marta Pienkowska (ET
 H Zürich); and Josep de la Puente and Rosa M. Badia (Barcelona Supercomput
 ing Center (BSC))\n---------------------\nWorkflows on LUMI: Europe's most
  powerful supercomputer\n\nTomasz Malkiewicz (CSC - IT Center for Science)
 \n---------------------\nWORKS24 — Afternoon Break\n---------------------\
 nEnabling Low-Overhead HT-HPC Workflows at Extreme Scale using GNU Paralle
 l\n\nGNU Parallel is a versatile and powerful tool for process paralleliza
 tion widely used in scientific computing. This paper demonstrates its effe
 ctive application in high-performance computing (HPC) environments, partic
 ularly focusing on its scalability and efficiency in executing large-scale
  high-thr...\n\n\nKetan Maheshwari (Oak Ridge National Laboratory (ORNL));
  William Arndt (Lawrence Berkeley National Laboratory (LBNL)); and Ahmad M
 aroof Karimi, Junqi Yin, Frederic Suter, Seth Johnson, and Rafael Ferreira
  da Silva (Oak Ridge National Laboratory (ORNL))\n---------------------\nT
 rust and Verification of AI-Based Decision Making for Future Scientific  W
 orkflows: Challenges and Solutions\n\n-\n\n\nAnna Giannakou, Oluwamayowa A
 musat, and Lavanya Ramakrishnan (Lawrence Berkeley National Laboratory (LB
 NL))\n---------------------\nWORKS24 — Morning Break\n\nTag: Applications 
 and Application Frameworks, Distributed Computing, Middleware and System S
 oftware\n\nRegistration Category: Workshop Reg Pass\n\nSession Chairs: Sil
 vina Caino-Lores (National Institute for Research in Digital Science and T
 echnology (Inria)) and Anirban Mandal
END:VEVENT
END:VCALENDAR