Established in 2003 with support from NASA and the National Science Foundation (NSF), the Vermont Advanced Computing Center (VACC) is a cornerstone of the University of Vermont’s research enterprise. As a premier high-performance computing facility in New England, the VACC serves as a multidisciplinary hub where advanced technology meets scientific discovery.

Our Mission

The VACC is dedicated to accelerating research impact through three strategic pillars:

• Advancing High-Impact Research: Supporting complex, data-intensive projects across the physical sciences, engineering, medicine, and the humanities.
• Promoting Computational Education: Empowering the next generation of innovators by integrating high-performance computing (HPC) into the classroom and providing hands-on technical outreach.
• Investing in Sustainable Infrastructure: Maintaining and evolving a state-of-the-art computational ecosystem—including our BlueMoon and IceCore clusters—to provide researchers with a competitive edge.

Empowering the Research Community

From modeling climate change and drug discovery to exploring high-throughput AI and machine learning workflows, the VACC provides the massive parallel processing power and specialized storage (such as the DataMountain database cluster) required to solve the most pressing challenges of our time. We are committed to providing UVM faculty, staff, and students with the secure, scalable, and expertly supported environment necessary to turn ambitious ideas into peer-reviewed results. For Cluster Specifications, https://www.uvm.edu/vacc/cluster-specs

The VACC provides the high-performance computing (HPC) resources and technical expertise essential for data-intensive research and discovery. We bridge the gap between complex data challenges and actionable insights.

Our Core Services

• Scalable High-Performance Computing: Access a robust computational environment engineered for speed and reliability. Our infrastructure is designed to handle massive parallel processing tasks while adhering to modern cybersecurity and institutional compliance standards.
• Strategic Data Management & Storage Secure your research throughout its lifecycle. We provide high-capacity storage solutions, including data curation, archiving, and publication tools, to meet grant requirements and promote scientific reproducibility.
• Curated Scientific Software Hit the ground running with an optimized software stack. We actively maintain an extensive library of scientific packages and libraries, ensuring compatibility and peak performance on our clusters.
• Empowering the Next Generation. We go beyond hardware. The VACC offers hands-on workshops, web-based training, and expert consultation to help faculty integrate advanced cyberinfrastructure into their research and classroom curricula.
• Custom Software & Interface Engineering: Leverage our team’s deep expertise in developing specialized software and web interfaces. We help you build and deploy applications specifically tuned for supercomputing environments.

In 2003, with early support from NASA and Senator Patrick Leahy, the University of Vermont (UVM) conducted a comprehensive campus-wide assessment of current needs and future directions for high-performance computing (HPC) within its research enterprise. This initiative included consultations with an expert panel from the American Association for the Advancement of Science (AAAS) and with UVM research leadership and faculty members from multiple disciplines.

Building on this foundation, the facility, commonly referred to as the "Bluemoon cluster," was primarily developed using IBM systems architecture. Since its inception, the cluster has undergone three major upgrades, incorporating next-generation IBM high-performance computing hardware to enhance performance and data storage, and to improve data security and energy efficiency for a broader user base.

To further advance its HPC capabilities, in 2019 the university introduced a new computing cluster, DeepGreen, funded by a $1 million grant from the National Science Foundation (NSF). This massively parallel cluster comprises over 70 GPUs and is capable of mixed-precision calculations exceeding 8 petaflops, using the NVIDIA Tesla V100 architecture. The hybrid design accelerates high-throughput artificial intelligence and machine learning workflows, and its extensive parallelism enables the development of new and transformative research pipelines.

The following year, 2020, UVM's inclusion in the AMD University Program underscored its role as a national leader in accelerating computing to drive innovation, leading to a $1 million award from the AMD HPC Fund. This partnership was instrumental in advancing critical research during the global pandemic and continues to support UVM's advanced modeling capabilities.

Continuing this momentum, in 2021, the NSF awarded the VACC a grant, “MRI: Acquisition of a Massive Database to Accelerate Data Science Discovery,” totaling $1,035,737, with joint support from the EPSCoR program, to construct DataMountain. This specialized database cluster addresses big data challenges by overcoming the performance limitations of traditional computing systems known as the memory wall.

This commitment to advancement was further demonstrated in 2023, when AMD contributed a $100,000 gift to support teaching and training on GPU systems at UVM's VACC.

VACC serves as a critical infrastructure partner for the state’s NSF EPSCoR RII Track-1 award, known as SOCKS (Science of Online Corpora, Knowledge, and Stories). The SOCKS project, a $20 million, five-year initiative (2023–2028), is designed to harness the data revolution by quantifying how stories and narratives influence social, economic, and health domains. A core component of this grant is the expansion of cyberinfrastructure to support large-scale analysis of datasets, such as Storywrangler, with a $1,200,000 investment in new cluster hardware.

Most recently, in 2025, a $2.1 million grant from the National Science Foundation funded the development of IceCore, a new artificial intelligence-focused cluster reported to be 100 times faster than previous UVM systems. https://www.uvm.edu/cems/news/icecore-advances-new-era-discovery-uvm