REX-IO '24: Workshop on Re-envisioning Extreme-Scale I/O for Emerging Hybrid HPC Workloads Kobe International Conference Center Kobe, Japan, September 24, 2024

High Performance Computing (HPC) applications are evolving to include not only traditional modeling and simulation bulk-synchronous scale-up workloads but also scale-out workloads, including artificial intelligence (AI), big data analytics methods, deep learning, and complex multi-step workflows. With the advent of Exascale systems such as Frontier, workflows include multiple different components from both scale-up and scale-out communities operating together to drive scientific discovery and innovation. With the often conflicting design choices between optimizing for write- vs. read-intensive, having flexible I/O systems is crucial to support hybrid workloads. Another performance aspect is the intensifying complexity of parallel file and storage systems in large-scale cluster environments. Storage system designs are advancing beyond the traditional two-tiered file system and archive model by introducing new tiers of temporary, fast storage close to the computing resources with distinctly different performance characteristics. 

The changing landscape of emerging hybrid HPC workloads along with the ever increasing gap between the compute and storage performance capabilities reinforces the need for an in-depth understanding of extreme-scale I/O and for rethinking existing data storage and management techniques. Traditional approaches of managing data might fail to address the challenges of extreme-scale hybrid workloads. Novel I/O optimization and management techniques integrating machine learning and AI algorithms, such as intelligent load balancing and I/O pattern prediction, are needed to ease the handling of the exponential growth of data as well as the complex hierarchies in the storage and file systems. Furthermore, user-friendly, transparent and innovative approaches are essential to adapt to the needs of different HPC I/O workloads while easing the scientific and commercial code development and efficiently utilizing extreme-scale parallel I/O and storage resources.

Established at IEEE Cluster 2021, the Re-envisioning Extreme-Scale I/O for Emerging Hybrid HPC Workloads (REX-IO) workshop has created a forum for experts, researchers, and engineers in the parallel I/O and storage, compute facility operation, and HPC application domains. REX-IO solicits novel work that characterizes I/O behavior and identifies the challenges in scientific data and storage management for emerging HPC workloads, introduces potential solutions to alleviate some of these challenges, and demonstrates the effectiveness of the proposed solutions to improve I/O performance for the exascale supercomputing era and beyond. We envision that this workshop will contribute to the community and further drive discussions between storage and I/O researchers, HPC application users and the data analytics community to give a better in-depth understanding of the impact on the storage and file systems induced by emerging HPC applications.

Submission Guidelines

All papers must be original and not simultaneously submitted to another journal or conference. Indicate all authors and affiliations. All papers will be peer-reviewed using a single-blind peer-review process by at least three members of the program committee. Submissions should be a complete manuscript. REX-IO accepts traditional research papers (page limit: 8 pages + 2 additional pages) for in-depth topics and short papers (page limit: 4 pages + 1 additional page) for work in progress on hot topics.

Paper format: single-spaced double-column pages using 10-point size font on 8.5x11 inch pages (IEEE conference style), including figures, tables, and references. The submitted manuscripts should include author names and affiliations. The IEEE conference style templates for MS Word and LaTeX provided by IEEE eXpress Conference Publishing are available here: https://www.ieee.org/conferences/publishing/templates.html 

Papers are to be submitted electronically in PDF format. Submitted papers should not have appeared in or be under consideration for a different workshop, conference or journal. All accepted papers need to be presented at the workshop by one of the authors.

All accepted papers (subject to post-review revisions) will be published in the IEEE Cluster 2024 companion proceedings.

Important Dates

Please note: All Dates are Anywhere on Earth

• Submissions open: June 1, 2024

• Submission deadline: July 10, 2024 July 17, 2024 (final extension)

• Notification to authors: July 26, 2024 July 31, 2024 (final extension)

• Author Registration due: August 2, 2024

• Camera-ready paper due: August 9, 2024

• Workshop date: September 24, 2024

List of Topics

• Understanding I/O inefficiencies in emerging workloads such as complex multi-step workflows, in-situ analysis, AI, and data analytics methods 

• New I/O optimization techniques, including how ML and AI algorithms might be adapted for intelligent load balancing and I/O pattern prediction of complex, hybrid application workloads

• Performance benchmarking and modeling, and I/O behavior studies of emerging workloads

• New possibilities for the I/O optimization of emerging application workloads and their I/O subsystems

• Efficient monitoring tools for metadata and storage hardware statistics at runtime, dynamic storage resource management, and I/O load balancing

• Parallel file systems, metadata management, and complex data management

• Understanding and efficiently utilizing complex storage hierarchies beyond the traditional two-tiered file system and archive model 

• User-friendly tools and techniques for managing data movement among compute and storage nodes

• Use of staging areas, such as burst buffers or other private or shared acceleration tiers for managing intermediate data between computation tasks

• Application of emerging big data frameworks towards scientific computing and analysis

• Alternative data storage models, including object and key-value stores, and scalable software architectures for data storage and archive

• Position papers on related topics

Committees

Program Committee

• Hadeel Albahar (Kuwait University, Kuwait)

• Tyler Allen (University of North Carolina at Charlotte, USA)

• Phil Carns (Argonne National Laboratory, USA)

• Hariharan Devarajan (Lawrence Livermore National Laboratory, USA)

• Anna Fuchs (University of Hamburg, Germany)

• Shadi Ibrahim (National Institute for Research in Digital Science and Technology (Inria), France)

• Ahmad Maroof Karimi (Oak Ridge National Laboratory, USA)

• Hideyuki Kawashima (Keio University, Japan)

• Michael Kuhn (Otto von Guericke University Magdeburg, Germany)

• Radita Liem (RWTH Aachen University, Germany)

• M. Mustafa Rafique (Rochester Institute of Technology, USA)

• Kento Sato (RIKEN R-CCS, Japan)

• Houjun Tang (Lawrence Berkeley National Laboratory, USA)

• Osamu Tatebe (University of Tsukuba, Japan)

• Mai Zheng (Iowa State University, USA)

Organizing committee

• Arnab K. Paul (BITS Pilani, K K Birla Goa Campus, India)

• Sarah M. Neuwirth (Johannes Gutenberg University Mainz, Germany)

• Jay Lofstead (Sandia National Laboratories, USA)

Contact

All questions about submissions should be emailed to <rexio24 AT easychair DOT org>