ISC 2026
- Hamburg, Germany
Stop by the Globus booth F11 at ISC 2026. Learn more about our latest services including Globus Compute and Globus Streams. Join the following sessions:
Date: June 22
Time: 9 AM -1 PM
Location: Hall X9-1st Floor
Speakers:
Kyle Chard, University of Chicago
Ian Foster, Argonne National Lab
Alok Kamatar, University of Chicago
Building Scalable Agentic Systems for Science: Concepts, Architectures, and Hands-On with Academy
Agentic systems, in which autonomous agents collaborate to solve complex problems, are emerging as a transformative methodology in AI. However, adapting agentic architectures to scientific cyberinfrastructure—spanning HPC systems, experimental facilities, and federated data repositories—introduces new technical challenges. In this half-day tutorial, we introduce participants to the design, deployment, and management of scalable agentic systems for scientific discovery. We will present Academy, a Python-based middleware platform built to support agentic workflows across heterogeneous research environments. Participants will learn core agentic system concepts, including asynchronous execution models, stateful agent orchestration, and dynamic resource management. A guided hands-on session will help attendees build and launch their own agentic workflows. We will present case studies in materials discovery, biology, and chemistry. This tutorial is designed for researchers, developers, and cyberinfrastructure professionals interested in advancing AI-driven science with next-generation autonomous systems.
Date: June 23, 2026
Time: 4:00 PM - 4:30 PM
Location: Hall Z - 3rd Floor
Speakers:
Kyle Chard, University of Chicago
Rafael Ferreria da Silva, Oak Ridge National Laboratory\
Autonomous Discovery at Exascale Through Agentic AI and Multi-Agent Coordination
Scientific discovery is entering a new phase in which AI systems no longer just accelerate individual steps of the research process but actively coordinate them. This talk examines how agentic AI and multi-agent systems can evolve scientific workflows from human-directed pipelines into autonomous loops of hypothesis generation, simulation, analysis, and refinement. Drawing on emerging patterns across HPC and AI-for-science efforts, I will discuss how coordinated agents can plan campaigns, allocate heterogeneous resources, and adapt to intermediate results, turning exascale capability into exascale productivity. I will also highlight the open challenges around reliability, reproducibility, and human oversight, and sketch a credible path toward autonomous discovery at exascale.
Date: June 25, 2026
Time: 11:25 - 11:45 AM
Location: Hall E - 2nd Floor
Speaker: Kyle Chard, University of Chicago
Icicle: Scalable Metadata Indexing and Real-Time Monitoring for HPC File Systems
Modern HPC file systems can contain billions of files and hundreds of petabytes of data, making even simple questions increasingly intractable to answer. Traditional file system utilities such as find and du fail to scale to these sizes. While external indexing tools like GUFI and Brindexer improve query performance, they remain batch-oriented and unsuitable for heterogeneous, rapidly evolving environments.
We present Icicle, a scalable framework for continuous file system metadata indexing and monitoring. Icicle maintains a unified, up-to-date, and queryable view of file system state while supporting both periodic snapshot-based ingestion for bulk metadata updates and event-based ingestion for real-time synchronization from production systems such as Lustre and IBM Storage Scale. Built on Apache Kafka and Apache Flink, Icicle provides high-throughput, fault-tolerant, and horizontally scalable ingestion of metadata events into two complementary search indexes, enabling both individual file discovery and aggregate summary statistics by user, group, and directory.
This architecture enables efficient support for both coarse-grained administrative queries and interactive analytics over billions of objects. Our experimental evaluation on production-scale HPC datasets demonstrates order-of-magnitude throughput improvements over existing monitoring and indexing approaches, with tunable options for balancing consistency, latency, and metadata freshness.