| 9:00am-9:05am |
PDSW 2026 Welcome
Organizers:
Jay Lofstead, Sandia National Laboratories
Sarah Neuwirth, Johannes Gutenberg University Mainz
Lipeng Wan, Georgia State University
Dean Hildebrand, Google LLC
Description
|
FEATURED TALK: Chair - Jay Lofstead
|
| 9:05am- 10:00am |
The Relationship Between AI and Research Data Management Christine Kirkpatrick (San Diego Supercomputer Center, USA)
Abstract
|
| 10:00am- 10:30am |
Morning Break
|
| MAIN SESSION PART 1: Chair - Ricardo Macedo |
| 10:30am-10:50am |
GPU-Aware h5py: A Pure-Python Extension for HDF5--GPU Data Movement
Lena Oden,FernUniversitaet in Hagen, Germany
Abstract
|
| 15:00am-11:10am |
GPUH5: A GPU-first I/O Library for Scientific Data
Joshua Fernandes, Illinois Institute of Technology, USA
Luke Logan, Illinois Institute of Technology, USA
Xian-He Sun, Illinois Institute of Technology, USA
Anthony Kougkas, Illinois Institute of Technology, USA
Abstract
|
| 11:10am-11:30pm |
Dynamic Planning for Scalable Data Movement in Accelerated Query Engines
Richard Zamora, NVIDIA, USA
Tom Augspurger, NVIDIA, USA
Peter Entschev, NVIDIA, USA
Mads Kristensen, NVIDIA, USA
Lawrence Mitchell, NVIDIA, USA
Matthew Murray, NVIDIA, USA
Niranda Perera, NVIDIA, USA
Vyas Ramasubramani, NVIDIA, USA
Matthew Roeschke, NVIDIA, USA
Benjamin Zaitlen, NVIDIA, USA
Abstract
|
| 11:30pm-11:50pm |
GDSight: Per-Operation Cross-Layer Attribution for GPUDirect Storage
Izzet Yildirim, Illinois Institute of Technology, USA
Xian-He Sun, Illinois Institute of Technology, USA
Anthony Kougkas, Illinois Institute of Technology, USA
Abstract
|
| 12:50pm-12:10pm |
Stress-Testing Exascale Scientific Data Output - ADIOS2 Configuration Strategies for WarpX under Extreme Load Imbalance
Greg Eisenhauer, Georgia Institute of Technology, USA
Junmin Gu, Lawrence Berkeley National Laboratory, USA
Axel Huebl, Lawrence Berkeley National Laboratory, USA
Scott Klasky, Oak Ridge National Laboratory, USA
Norbert Podhorszki, Oak Ridge National Laboratory, USA
Alexander Sim, Lawrence Berkeley National Laboratory, USA
Jean-Luv Vay, Lawrence Berkeley National Laboratory, USA
Kesheng (John) Wu, Lawrence Berkeley National Laboratory & ESnet, USA
Abstract
|
| 12:10pm-12:15pm |
[WIP]: Correctness Evidence for I/O Optimizations in VPIC-Kokkos: Preliminary Results
Connor Browne,University of Tennessee, Knoxville, and University of Washington, Bothell, USA
Nigel Tan and Scott Luedtke,Los Alamos National Laboratory, USA
Michela Taufer,University of Tennessee, Knoxville, USA
Brian Albright,Los Alamos National Laboratory, USA
Abstract | Paper
|
| 12:15pm-12:20pm |
[WIP]: Trigger-Gated Multimodal LLM Agents for In-situ Scientific Simulation Monitoring
Hua Xu, Jaime Cernuda, Illinois Institute of Technology, USA
Anthony Kougkas, Illinois Institute of Technology, USA
Ana Gainaru, Oak Ridge National Laboratory, USA
Scott Klasky, Oak Ridge National Laboratory, USA
Xian-He Sun, Illinois Institute of Technology, USA
Jay Lofstead, Sandia National Laboratories and Univ. of New Mexico, USA
Abstract | Paper
|
| 12:20pm-12:25pm |
[WIP]: Fair, Random IO Operations are Hard
Jay Lofstead, Sandia National Laboratories and Univ. of New Mexico, USA
Dean Hildebrand, Google LLC, USA
Andreas Dilger, The Lustre Collective
Jean Luca Bez, ESnet / Lawrence Berkeley National Laboratory, USA
Julian Kunkel, GWDG, Germany
Abstract | Paper
|
| 12:25pm-12:30pm |
[WIP]: One Bus Accident Away from Death: The Fragility of HPC I/O Tools
Ray A. O. Sinurat (University of Chicago, USA)
Radita Liem, Johannes Gutenberg University Mainz, Germany
Sarah Neuwirth, Johannes Gutenberg University Mainz, Germany
Abstract | Paper
|
| 12:30pm-2:00pm |
Lunch Break
|
| MAIN SESSION PART 2: Chair - TBD |
| 2:00pm-2:20pm |
Sharing a Fabric with Collective Communication: Two Storage Penalties in DL Training
Chen Wang, Nanyang Technological University, Singapore
Wenzhao Wu, Nanyang Technological University, Singapore
Hyojin Kim, Lawrence Livermore National Laboratory, USA
Jae-Seung Yeom, Lawrence Livermore National Laboratory, USA
Abstract
|
| 2:20pm-2:40pm |
Copy2: Data Sync Tool For High Performance Storages
Aiden Lambert, Georgia Institute of Technology, USA
Aaron Jezghani, Georgia Institute of Technology, USA
Deepa Phanish, Georgia Institute of Technology, USA
Abstract
|
| 2:40pm-3:00pm |
Measurement-Driven Contention-Aware, Herd-Safe Stripe Placement for Lustre
Zhaobin Zhu, Johannes Gutenberg University Mainz, Germany
Sarah Neuwirth, Johannes Gutenberg University Mainz, Germany
Abstract
|
| 3:00pm-3:30pm |
Afternoon Break
|
| MAIN SESSION PART 3: Chair - Sarah Neuwirth |
| 3:30pm-3:50pm |
STARCH: RPC-Lifecycle Tracing of Margo-Based Distributed Systems for Performance Analysis
Alexander Klassen, University of Tsukuba, Japan
Osamu Tatebe, University of Tsukuba, Japan
Abstract
|
| 3:50pm-4:10pm |
PRISM: Lightweight Surrogation for Lossy Compression Quality Prediction
Adelina Zheng, Brown University, USA
Dong Dai, University of Delaware, USA
Sheng Di, Argonne National Laboratory, USA
Franck Cappello, Argonne National Laboratory, USA
Jinzhen Wang, University of North Carolina at Charlotte, USA
Abstract
|
| 4:10pm-4:30pm |
Large Scale Sorting on UPMEM Processing-in-Memory Architectures
Matthew Gardner, Northern Arizona University, USA
James Harris, Northern Arizona University, USA
Revanth Munugala, Northern Arizona University, USA
Tasmia JannatMissouri University of Science and Technology, USA
Satish Puri, Missouri University of Science and Technology, USA
Michael Gowanlock, Northern Arizona University, USA
Abstract
|
| 4:30pm-4:35pm |
[WIP]: DAOS Update: Version 2.8 and MLPerf Storage
Michael Hennecke, Hewlett Packard Enterprise and DAOS Foundation
Abstract | Paper
|
| 4:35pm-4:40pm |
[WIP]: Exploring DAOS as a Burst Buffer for Nx100 Gbps Scientific Streaming
Xinxin Mei, Thomas Jefferson National Accelerator Facility, USA
Jie Ren, College of William & Mary, USA
Jon Zarling, Thomas Jefferson National Accelerator Facility, USA
Cameron Clarke, Thomas Jefferson National Accelerator Facility, USA
Vardan Gyurjyan, Thomas Jefferson National Accelerator Facility, USA
Abstract | Paper
|
| 4:40pm-4:45pm |
[WIP]: Move Metadata, Not Data: A Register-in-Place Lakehouse for Large Autonomous-Driving Datasets
Jey Kang, Gwangju Institute of Science and Technology, South Korea
Kangryeol Kim, Gwangju Institute of Science and Technology, South Korea
JinWang Mok, Gwangju Institute of Science and Technology, South Korea
Ho Kim, Gwangju Institute of Science and Technology, South Korea
Hyunje Sung, Gwangju Institute of Science and Technology, South Korea
Sun Park, Gwangju Institute of Science and Technology, South Korea
JongWon Kim, Gwangju Institute of Science and Technology, South Korea
Abstract | Paper
|
| 4:45pm-4:50pm |
[WIP]: Type-Safe, Dynamically Defined Metadata Support for Scientific Applications
Joshua Teves, Sandia National Laboratories, USA
Curtis Brinker, Sandia National Laboratories, USA
Anthony Kougkas, Illinois Institute of Technology, USA
Ana Gainaru, Oak Ridge National Laboratory, USA
Scott Klasky, Oak Ridge National Laboratory, USA
Xian-He Sun, Illinois Institute of Technology, USA
Jay Lofstead, Sandia National Laboratories and Univ. of New Mexico, USA
Abstract | Paper
|
| 4:50pm-5:15pm |
Invited Talk: AI Agents for Data Infrastructure: From Telemetry to Optimization
Dong Dai, University of Delaware, USA
Abstract
|
| 5:15pm-5:20pm |
[WIP]: GPU-Initiated Data Movement for Accelerator-Centric HPC
Baodi Shan, Stony Brook University, USA
Mauricio Araya-Polo, TotalEnergies, France
Barbara Chapman, Stony Brook University, USA
Abstract | Paper
|
| 5:20pm-5:25pm |
[WIP]: DPUCrumbs: Dissecting DPU Pathways
Ray A. O. Sinurat, University of Chicago, USA
Sarah Neuwirth, Johannes Gutenberg University Mainz, Germany
Radita Liem, Johannes Gutenberg University Mainz, Germany
Jay Lofstead, Sandia National Laboratories, USA
Nikoli Dryden, Lawrence Livermore National Laboratory, USA
Hariharan Devarajan, Lawrence Livermore National Laboratory, USA
Abstract | Paper
|
| 5:25pm-5:30pm |
[WIP]: Towards PIM-Aware Expert Decomposition and Mapping for MoE Inference
Sitian Chen, Hong Kong Baptist University, Hong Kong
Haonan Wang, Hong Kong Baptist University, Hong Kong
Amelie Chi Zhou, Hong Kong Baptist University, Hong Kong
Abstract | Paper
|
The increasing importance of efficient data storage and management continues to drive scientific productivity across traditional simulation-based HPC environments and emerging Cloud, AI/ML, and Big Data analysis frameworks. Challenges are compounded by the rapidly expanding volumes of experimental and observational data, the growing disparity between computational and storage hardware performance, and the rise of novel data-driven algorithms in machine learning. This workshop aims to advance research and development by addressing the most pressing challenges in large-scale data storage and processing.
We invite the community to contribute original research manuscripts that introduce and evaluate novel algorithms or architectures, share significant scientific case studies or workloads, or assess the reproducibility of previously published work. We emphasize the importance of community collaboration for problem identification, workload capture, solution interoperability, standardization, and shared tools. Authors are encouraged to provide comprehensive experimental environment details (software versions, benchmark configurations, etc.) to promote transparency and facilitate collaborative progress.
All submissions to the PDSW’25 will undergo a rigorous double-anonymous peer review process overseen by the workshop program committee. Successful submissions will be published in the SC25 Workshop Proceedings and featured on the workshop website alongside associated talk slides.
WIP papers will not be included in the conference proceedings but will be made available through an archival service after the event. To participate, authors must submit a one-page abstract using the IEEE conference paper template.
