A New Computational Powerhouse for New Mexico State University
New Mexico State University launched Colibrí this summer. This high-performance computing system significantly expands the research capacity of the institution by providing local control over artificial intelligence, machine learning, and large-scale data analysis. The system serves as a central hub for researchers to build and test models before moving them into real-world applications. Its deployment addresses specific regional challenges that demand heavy computational lifting.
The system hardware is extensive. Colibrí contains eight high-density CPU compute nodes alongside five GPU nodes. These include 24 NVIDIA RTX 6000 Blackwell GPUs and eight NVIDIA H200 GPUs. A 7-petabyte VDURA parallel-file-system storage appliance backs the entire setup. This combination allows for both traditional scientific simulations and the data-intensive training required for modern artificial intelligence tools.
Solving Local Problems with Advanced Modeling
The university intends to apply this power to issues directly affecting New Mexico. Researchers are currently looking at wildfire detection, which requires the rapid analysis of sensor data to coordinate emergency responses. The system will also host digital twins of agricultural networks and water treatment plants. By creating these virtual replicas, scientists can test different variables and outcomes in a safe, controlled environment before deploying adjustments to the actual physical infrastructure.
Remote energy systems represent another primary area of focus. These projects often involve hybrid solar, wind, and fuel-cell setups located in areas with limited power or connectivity. Engineers develop the foundational models on Colibrí and then translate them into edge decision-support tools. This ensures the technology remains effective even when disconnected from a central grid. These applications highlight the practical shift from theoretical study to functional, on-the-ground solution development.
Financial Backing and Statewide Impact
The initiative is the result of years of planning by the NMSU Research Cores Program. Financial support arrived from multiple channels. The New Mexico IDeA Networks of Biomedical Research Excellence contributed nearly $1 million specifically for hardware. This funding ensures that researchers and students across 10 different higher education institutions in the state gain access to the system. Shelley Lusetti, director of NM-INBRE, stated that this investment broadens the reach of computational tools for biomedical data science throughout the state.
Additionally, the National Science Foundation awarded $1.4 million to a project led by Lizely Madrigal, the NMSU director of University Technology Solutions and Support. These funds cover the data-storage system and the necessary personnel to manage the infrastructure. The project is already attracting external partnerships as well. NMSU is starting a collaboration with the Fujitsu corporation to bring new energy-efficient chip technology to the campus this fall.
Building a Lasting Research Ecosystem
The university views Colibrí as more than just a cluster of servers. It is a catalyst for a larger research ecosystem. By bringing together experts from the main campus, the Physical Science Laboratory, and external bodies like Los Alamos National Laboratory and the San Diego Supercomputer Center, NMSU aims to create a hub of shared knowledge. Amy Wagler, the director of research computing and data science, emphasized that this resource is designed for solution-oriented problem-solving.
The work is far from finished. With the hardware now online, the focus shifts to training students and faculty on how to use these tools effectively. The goal is to make research computing a permanent, reliable asset for New Mexico. This approach aligns with the mission of a land-grant university, ensuring the benefits of high-level academic research reach beyond the classroom and into the state economy.