Vanderbilt Advanced Computing Center for Research and Education Selects Hammerspace to Power Next-Generation Research Data Infrastructure

Hammerspace, the high-performance data platform for AI Anywhere, announced that the Vanderbilt Advanced Computing Center for Research and Education (ACCRE) at Vanderbilt University has selected Hammerspace to modernize its research data infrastructure.

ACCRE, Vanderbilt’s campus-wide HPC resource and research support facility, provides advanced computing and storage services for faculty and students across disciplines ranging from genetics and physics to engineering and social sciences. With a mission to “explore and benefit from the new world of computing,” ACCRE enables researchers to run large-scale simulations, data analyses and ML models critical to advancing discovery.

To meet growing data demands across hundreds of research projects, ACCRE sought a more flexible and cost-efficient approach to managing petabytes of research data. Historically, ACCRE has operated separate systems for primary and archive storage, including Panasas, GPFS and LStore. ACCRE wanted a solution that could unify its diverse storage tiers, leverage commodity hardware and dynamically provision storage resources across compute and GPU nodes.

After evaluating many vendors, ACCRE selected Hammerspace to deploy a 10-petabyte environment integrating CPU/GPU server-local storage for Tier 0 performance, newly purchased commodity storage servers for tier-1, and multi-petabyte archival capacity from its existing LStore environment, all under a single global namespace.

Hammerspace powers next-generation research data infrastructure
at Vanderbilt Advanced Computing Center for Research and Education (ACCRE)

By adopting Hammerspace in combination with LStore, ACCRE expects to reduce its average cost of storage by 48% while providing faster, more flexible data access to the Vanderbilt research community. The Hammerspace Data Platform’s open architecture aligns with LStore’s key characteristics to use commodity hardware instead of proprietary storage appliances, improving flexibility and reducing vendor lock-in.

“Hammerspace’s platform offers a composable, open architecture that lets us unify GPU server-local, tiered and archival storage into a single data environment,” said Hunter Hagewood, executive director, research computing operations, ACCRE. “Combined with LStore, our largest and most advanced storage platform, we now have a long-term strategy for meeting strong capacity demand. This integration not only cuts our costs dramatically but also changes how we deliver compute and storage services to Vanderbilt researchers, making it easier to support the next gen of data-driven science.”

With approximately 750 compute nodes and 80 GPU nodes, ACCRE’s HPC environment supports workloads in AI, data analytics and simulation-based research. By integrating Hammerspace’s Data Platform, the Center gains the ability to quickly compose and reconfigure storage environments based on project needs, improving collaboration and data throughput for multidisciplinary research teams.

“Vanderbilt ACCRE is a perfect example of how Hammerspace empowers research institutions to modernize their data infrastructure,” said Jeff Giannetti, CRO, Hammerspace. “By unifying performance, capacity and archive storage into one high-performance data platform, ACCRE can deliver faster insights, greater efficiency and new agility to its scientific community.”