Enabling integration across research infrastructures for accelerating science
March 06, 2025 | Ryan Chard
Welcome to our blog series on using Globus to integrate diverse research infrastructures to accelerate science. In today’s data-intensive scientific landscape, it is increasingly necessary to make use of resources that are located at different facilities. For example, as data may be acquired in one location, analyzed in another, and published in yet another location. DOE’s Integrated Research Infrastructure program (https://iri.science/) exemplifies these new requirements with an aim to “link experimental and observational scientific user facilities, data assets, and advanced computing resources so that researchers can combine these tools in novel ways that radically accelerate discovery.”
The Globus platform provides a robust, scalable, secure, and globally accessible cyberinfrastructure that removes the frictions associated with managing research processes that span locations. Globus enables researchers to easily access, use, and combine distributed research infrastructure, such as High-Performance Computing (HPC) clusters, cloud-hosted storage, and scientific instruments—no matter where they are located.
Over the course of this series we will showcase six real-world applications that highlight how Globus has enabled production scientific workflows to integrate diverse infrastructure–from real-time analysis of data from synchrotron light sources to on-demand visualization and processing of enormous cosmology datasets via a web-based data portal, these use cases highlight the versatility and efficiency of Globus in facilitating cross-facility collaboration.
Join us as we explore how Globus empowers researchers to break down barriers, streamline processes, and drive innovation across a wide array of scientific domains. Whether you’re a seasoned researcher or new to the world of integrated research infrastructure, this series will provide valuable insights into the future of collaborative science. Stay tuned for an exciting journey into the heart of modern scientific discovery!
Resilient X-ray Photon Correlation Spectroscopy With Facility Failover
The recently upgraded Advanced Photon Source (APS) is set to revolutionize X-ray science, enabling unprecedented levels of brightness and coherence for probing materials at the atomic scale. However, with great power comes great responsibility—specifically, the need for reliable, on-demand high-performance computing (HPC) resources to handle the massive data collection rates and dynamically guide experiments in real time. To address this challenge, researchers at Argonne National Laboratory have developed a scalable, resilient and automated approach for X-ray Photon Correlation Spectroscopy analysis, leveraging the power of multi-site computing, streamlined failover mechanisms, and Globus.
X-ray Photon Correlation Spectroscopy (XPCS) is a non-invasive technique that is used to investigate the dynamics of materials on the nanoscale. Specifically, XPCS studies the time-dependent fluctuations in the scattering of x-rays from a sample. These fluctuations can provide valuable information on the motion and interactions of atoms and molecules within the material, as well as the physical processes that underlie these phenomena. XPCS has a variety of potential applications in materials science, chemistry, and physics, where it can be used to study a wide range of materials, including liquids, glasses, and some types of crystals.
The Globus-based approach uses Globus Flows to automate the processing of XPCS data as it is collected. These flows are designed to handle the analysis process, from data transfer to analysis and publication—ensuring that experiments can proceed without manual intervention. By integrating Globus Flows in the XPCS analysis process , researchers can spend more time focusing on their science while Globus handles the laborious tasks of dealing with distributed research infrastructure. Globus Compute abstracts the differences between computing resources, enabling seamless multi-site computing. This means that computational tasks can be distributed across different HPC facilities, such as the Argonne Leadership Computing Facility (ALCF) and the National Energy Research Scientific Computing Center (NERSC), without requiring researchers to deal with the complexities of each system. Globus Transfer ensures that data is efficiently moved between facilities, minimizing latency and maximizing throughput. One of the most innovative aspects of this approach is that the Globus transfer and compute endpoints to be used are dynamically selected at invocation time. This allows the beamline to redirect the flow to move data and analysis tasks to an alternative facility if the primary facility becomes unavailable. For example, if ALCF experiences an outage, the XPCS workflow can continue analyses at NERSC, ensuring that experiments continue without interruption.
Contributors: APS: Hannah Parraga, Suresh Narayanan, Miaoqi Chu, Nicholas Schwarz; Globus: Labs/Globus: Rafael Vescovi, Nick Saint, Ryan Chard, Ben Blaiszik