University of Nebraska Deploys NSF-Funded ‘PLUMAGE’ GPU Infrastructure for AI Research

Key Takeaways

• PLUMAGE adds six NVIDIA H200 and 52 L40S GPUs to Nebraska’s HPC portfolio
• Researchers can run jobs via traditional batch scheduling or dynamic Kubernetes
• Free access extends to external collaborators, expanding the user community
• NSF funding emphasizes the novelty of linking local and national resources

Pulse Analysis

The rollout of PLUMAGE marks a significant upgrade to the University of Nebraska’s high‑performance computing (HPC) capabilities. Backed by a $700,000 National Science Foundation award, the platform bundles some of NVIDIA’s most powerful AI accelerators—six H200 and 52 L40S GPUs—into a flexible environment. By coupling the existing Swan cluster’s batch scheduler with the NSF‑run National Research Platform’s Kubernetes orchestration, PLUMAGE lets scientists select the execution model that best fits their workload, whether it’s a large‑scale training run or an interactive experiment.

Beyond raw compute power, PLUMAGE is designed to democratize AI research across the university’s diverse academic units. Faculty in business, architecture, social sciences and other traditionally low‑compute fields can now tap into GPU‑driven machine‑learning pipelines without needing dedicated hardware or deep systems expertise. The center’s staff provide tailored training and support, addressing the “intimidation factor” often associated with supercomputing. This outreach is expected to spark a wave of interdisciplinary projects, from predictive analytics in economics to generative design in architecture, accelerating the university’s overall research output.

Nationally, PLUMAGE exemplifies a growing trend toward hybrid, federated HPC infrastructures that blend local clusters with broader cyber‑infrastructure networks. By bridging Nebraska’s resources with the NSF’s National Research Platform, the system enhances data mobility and workload elasticity, positioning the university as a regional hub in the U.S. research ecosystem. As AI workloads continue to outpace traditional CPU‑centric models, similar GPU‑centric deployments are likely to become a cornerstone of academic and government research strategies, driving competitiveness and innovation across the country.