ON.energy Completes Medium-Voltage AI UPS Deployment at National Laboratory of the Rockies
ON.energy today announced the successful completion of its AI UPS™, marking a major milestone as the first medium-voltage uninterruptible power system specifically engineered for large-scale AI data centers. The system has been installed at the National Laboratory of the Rockies (NLR), providing a dedicated testing platform at the lab’s Flatirons Campus near Boulder, Colorado. This facility is home to one of the world’s most advanced grid simulators and hosts the Advanced Research on Integrated Energy Systems (ARIES) research platform, positioning it as an ideal environment for pioneering next-generation data center power solutions.
As AI campuses scale rapidly toward multi-gigawatt footprints, operators are encountering unprecedented challenges in power management. Modern AI compute environments, dominated by massive GPU arrays, generate steep power transients and rapid load swings. These fluctuations strain conventional uninterruptible power supply (UPS) architectures, which were originally designed for relatively steady IT loads. Consequently, data center operators are increasingly seeking solutions capable of maintaining voltage stability, delivering rapid ride-through performance during disturbances, and supporting both grid-connected and off-grid operations. ON.energy’s AI UPS deployment at NLR directly addresses this pressing industry need.
The collaboration between ON.energy and NLR is designed to rigorously test and validate the AI UPS under conditions that closely mimic real-world AI workloads. NLR will simulate complex power profiles characteristic of high-performance GPU clusters, including fast-moving computational spikes, dynamic load shifts, and potential fault conditions. The facility’s comprehensive grid simulation capabilities allow for the replication of a wide range of operating scenarios, including extreme voltage dips, frequency excursions, and black-start conditions. Through this testing, the AI UPS will demonstrate its ability to stabilize high-voltage loads, protect sensitive AI equipment, and mitigate the impact of disturbances on the surrounding electric grid.
In addition to testing system performance, the Flatirons Campus installation provides a controlled environment for prospective customers to conduct their own validation exercises. Hyperscale cloud providers and AI infrastructure operators have already expressed strong interest in utilizing this facility to model their own GPU workloads and grid integration scenarios. This approach enables data center operators to evaluate UPS performance for their specific operational requirements before committing to deployment at scale.
“Customers need to see how a medium-voltage UPS functions under real-world grid disturbances and the fast-moving load patterns of large AI deployments,” said Dax Kepshire, President of ON.energy’s Data Center division. “Our installation at NLR’s Flatirons Campus allows us to provide extensive performance validation in a controlled setting, ensuring our AI UPS meets the unique requirements of each customer’s workload and operational profile. It’s a critical step toward demonstrating how AI campuses can safely scale without overloading the grid or compromising compute performance.”
The AI UPS deployment at NLR also reflects broader trends in the energy and data center sectors. As AI compute demand continues to grow, the stress on the electric grid is intensifying. Rapid, high-amplitude power fluctuations from GPU clusters can introduce instability into local and regional grids, creating both operational and regulatory challenges. Medium-voltage UPS systems like ON.energy’s offer a solution by buffering these transients, providing ride-through during disturbances, and allowing AI campuses to operate in a grid-friendly manner. By integrating directly with the grid at medium voltage, these systems reduce the need for additional infrastructure upgrades while delivering resilience at scale.
“NLR is committed to advancing technologies that enhance the flexibility and reliability of the nation’s power systems at the pace this sector demands,” said Andrew Hudgins, ARIES Laboratory Program Manager. “Working with ON.energy, we have moved from first contact through contracting and construction in less than six months. The AI UPS deployment at Flatirons Campus will serve as a benchmark for how emerging energy solutions can integrate seamlessly with high-demand computing environments.”
For ON.energy, the Flatirons installation represents a critical step in validating its AI UPS as the backbone of grid-safe AI campuses. The system’s capabilities—rapid response to GPU load swings, voltage ride-through support, and grid stabilization—position it as a foundational technology for next-generation AI data centers. As hyperscalers, cloud providers, and AI infrastructure operators continue to scale their facilities, the ability to manage extreme power dynamics without compromising grid reliability is becoming a defining factor in campus design and operational planning.
Looking ahead, ON.energy plans to leverage the Flatirons Campus deployment as a demonstration platform for additional innovations in AI-focused power systems. By providing real-world performance data, controlled testing opportunities, and validation for customer-specific GPU workloads, the installation serves as both a research facility and a market-facing proof of concept. The partnership with NLR underscores a shared commitment to advancing the reliability, flexibility, and sustainability of AI infrastructure at a time when global demand for compute is accelerating at an unprecedented rate.
The successful deployment of the AI UPS also signals a shift in how the industry approaches medium-voltage power solutions for high-density AI campuses. Traditional UPS systems often struggle with the speed and magnitude of AI load fluctuations, requiring oversizing or additional infrastructure to mitigate risk. By contrast, ON.energy’s AI UPS is designed to integrate seamlessly into the campus power architecture, delivering rapid response to transient events, improved ride-through capabilities, and reduced operational stress on the grid. These features are essential for organizations looking to scale AI compute safely while maintaining resilience and efficiency.
As AI campuses expand toward multi-gigawatt operations, the need for innovative power infrastructure solutions becomes more urgent. ON.energy’s deployment at the National Laboratory of the Rockies demonstrates a forward-looking approach to meeting these challenges, combining advanced engineering with rigorous testing and validation. By addressing the unique demands of AI workloads, protecting the grid, and enabling customer-specific testing, the AI UPS positions ON.energy as a leader in the emerging field of grid-safe AI campus power systems.
With this milestone, ON.energy has not only proven the technical capabilities of its medium-voltage AI UPS but has also established a model for future collaborations between energy innovators and national laboratories. The Flatirons Campus deployment highlights the importance of bridging cutting-edge AI infrastructure with resilient, grid-friendly power solutions—a critical step toward enabling the next generation of high-performance computing while safeguarding electric system stability.
Source Link: https://www.businesswire.com/
