RMACC 2026

Beyond growing data volumes, researchers are now having to deal with increasing data velocity. For example, instruments are generating data faster than can be consumed by downstream processes and AI/ML-guided research is relying on near real-time feedback for optimizing experiments. Such data streaming applications require different tools for data management and computation than those used for traditional file-based data.

Building on the established and widely used Globus file transfer service, we present newly released capabilities that enable researchers to stream data securely across wide area networks in support of real-time data processing. We will provide an overview of the Globus data streaming service and describe how to incorporate it into your research applications.

Join us for a student career panel where you will hear from HPC professionals about how they entered the HPC workforce and what they are looking for when hiring new employees. This will be a one-hour session with presentations and time for questions

"HPC is a hidden gem that many students don’t even know exists - until they do." -- Panelist Kelly Byrne, Idaho National Labs

Flash pricing volatility has reshaped the storage landscape and organizations are rethinking how to get all-flash performance without runaway costs. In this session we will discuss practical approaches to navigating flash economics, cost optimization, and performance at scale.

NAND flash and DRAM prices are surging and supply is constrained, forcing requotes and delaying deals. VDURA’s mixed fleet with intelligent tiering architecture keeps AI/HPC projects moving with more budget stability. Attendees will learn how to:

• Keep AI/HPC projects on budget.
• Sustain throughput under concurrency and checkpointing.
• Avoid over-investing in flash for capacity-heavy data.

In short, VDURA’s mixed-fleet architecture delivers GPU-class performance with lower cost and greater supply-chain resilience.