ALBANY, NY, October 11, 2024 — The University at Albany has unveiled one of the most advanced AI supercomputers at a U.S. university as officials celebrate a major leap forward in New York’s access to world-changing technology. .
UAlbany’s new AI supercomputer is equipped with 24 of these NVIDIA DGX systems, for a total of 192 graphics processing units (GPUs). Photo credit: Brian Buescher.
Governor Cathy Hochul and State University of New York President John B. King Jr., along with University at Albany President Javidan Rodriguez, will discuss research in which scientists use a powerful new system to find treatments for devastating diseases. I attended a presentation at the RNA Research Institute at the University at Albany, one of the centers.
Powered by the NVIDIA DGX system, the new $16.5 million supercomputer is now the most capable AI supercomputer within SUNY and is a central pillar of UAlbany’s AI Plus initiative. This is a comprehensive push to integrate teaching and learning about AI across all universities, schools, and majors so that all graduates are prepared to live and work in a world fundamentally transformed by technology. .
“This new AI supercomputer is a huge step forward for the University at Albany for our faculty, staff, researchers, and especially our students,” Rodriguez said. “UAlbany is recognized as a national leader for our thoughtful and holistic approach to integrating AI into education and research, and this new AI supercomputer is the next important evolution in that strategy.
Associate Professor Alan Chen from the Department of Chemistry will demonstrate how researchers can use AI to create incredibly small structures that help filter toxins such as PFAS from water. Photo credit: Darren McGee/Office of Governor Kathy Hochul.
“Even more important than the technology is how we use it to help people,” Rodriguez said. “At UAlbany, we are focused on using AI for public impact research. Our researchers use this AI supercomputer to research atmospheric and climate science, public health, RNA, and nanotechnology. Advancing research in a variety of areas, including helping to build more resilient, healthier, and more equitable communities. That’s research that impacts the public good, and that’s what technology is really about. Value is where UAlbany shines.”
Empire AI preview
UAlbany’s new AI system is a key piece of what will happen under Gov. Kathy Hochul’s $400 million Empire AI initiative, an ambitious public-private consortium dedicated to making New York a leader in AI technology. It is also a preview. SUNY is a core partner in the Empire AI consortium, which is funding the construction of a new AI supercomputer in Buffalo. UAlbany’s system also provides SUNY campuses with critical access to advanced AI computing.
UAlbany’s AI Plus initiative took shape in February 2022, nine months before ChatGPT entered the public consciousness and the proliferation of AI became a topic of near-daily news coverage.
When I first arrived at the University at Albany, then Vice President for Research and Economic Development, I quickly realized that researchers were hungry for the computing power they needed to take their research to the next level. Ta.
“Our faculty’s ability to do the kind of research they wanted was limited by the computing power they had,” Kesavadas says. “To address real-world problems, we needed to invest in computing comparable to what was available in the private sector. So we decided to build a second-to-none computing system in New York. And most importantly, we want to empower teachers to think big and solve real-world problems.”
In April 2022, Governor Hochul and the state Legislature awarded $75 million to UAlbany. Half of that amount will go toward completing the university’s new downtown engineering building, now known as CNSE Downtown, and the other half will go toward expanding UAlbany’s AI supercomputing capabilities.
Shortly after, the University at Albany invested another stream of state funds in its largest cluster recruitment in its history. This is an exploration of 27 new faculty across all nine schools and colleges, delivering on the promise that AI Plus will be a comprehensive, university-wide approach. Along with many of UAlbany’s existing AI experts, these faculty members form the core of UAlbany’s new AI Plus Institute, which will accelerate basic and use case research in AI and bring together AI experts to research AI-related topics. Helps you deal with problems and questions.
Last month, the University at Albany appointed Prabha (Balakrishnan) Prabhakaran, a prominent AI scholar at the University of Texas at Dallas, as the institute’s first executive director.
“Companies are pushing the boundaries of AI with investments, but their scope is typically limited to what can generate revenue,” Kesavadas said. “Often the issues that really need attention aren’t at the top of the priority list unless industry decides it will lead to a future market. But as a public university, we have the ability to We have a responsibility and interest in addressing these issues.”
At Thursday’s presentation, Alan Chen, associate professor in the Department of Chemistry, and Sweta Vangaveti, a research scientist at the RNA Institute, said researchers will use supercomputers to test new structures and combinations to improve myotonic dystrophy. Demonstrated methods to develop treatments, antiviral drugs and filters for the novel coronavirus disease (COVID-19). Removes toxic PFAS chemicals from water.
“Drastically different” — What makes a computer an AI supercomputer?
Powered by 24 NVIDIA DGX systems, UAlbany’s new AI supercomputer is one of the largest at a U.S. university. The system is installed in a server cabinet in the university’s data center, has 192 NVIDIA Tensor Core GPUs, and is networked with approximately 3 miles of fiber-optic cable.
The accelerated computing power of GPUs makes them well-suited for high-intensity AI tasks that require massively parallel computations with large datasets to get answers quickly.
The interconnection of these powerful GPUs and their collective ability to quickly complete large amounts of complex calculations enables applications such as deep learning, a type of machine learning that can generate AI models for use cases such as computer vision. can be an effective research tool. , cancer and tumor prediction, and advanced weather forecasting. Calculations that normally took weeks with traditional systems can now be completed in days, and tasks that used to take days can now be completed in hours.
“We’ve always used high-performance computing clusters, but this is dramatically different,” said Brian Heaton, chief information officer at UAlbany, whose ITS staff helped design and build the system. I am. “The GPU is what makes a supercomputer an AI supercomputer, and our new AI system is designed for performance.”
“AI requires sophisticated mathematical calculations,” Heaton says. “The average computer cannot maintain the pace of computation that increases productivity within a reasonable time frame. Researchers developing AI often wait weeks for results. The more capabilities we can provide to researchers and faculty, the faster their response times will be and the more productive they will be in the field of AI.”
Hardware is only half the battle
Universities aren’t just investing in cutting-edge hardware. In addition to hiring 27 new faculty members specializing in AI (the largest hire in the university’s history), UAlbany is building a new ITS and research support staff.
CIO Brian Heaton (left) and the ITS project team helped design and build the new AI supercomputer and will continue to help teachers and researchers get the most out of it. Photo credit: Brian Buescher.
These new hires are supported by new SUNY funding dedicated to expanding the university’s research portfolio and include infrastructure automation engineers and data storage engineers who manage and maintain systems. Research technology analysts who work directly with faculty who are new to AI computing. AI Developer Analysts work with faculty who require high-level support.
“We needed to build a system that researchers at all levels could use without having to run into technical problems,” Heaton says. “But on the flip side, there are a lot of researchers who have been working on AI for a long time and need command-line access to their systems. We cover the whole spectrum.”
AI plus academic
At the heart of the AI Plus model is a new three-tier coursework progression available to all students in all majors. This advancement allows students to move from a general understanding of the basic principles of AI to a focus on AI-specific areas in their chosen field, such as the new Artificial Intelligence and Social Justice course in the School of Social Work.
These new offerings, such as the New Graduate Certificate in Data Science for Public Affairs at the Rockefeller School of Public Affairs and Policy, are in addition to more traditionally STEM-focused programs, such as the new AI concentration program in the School of Nanotechnology Science and Engineering. provided. Master’s degree in Computer Science or a new undergraduate minor in Machine Learning and Game Design in the College of Emergency Preparedness, Homeland Security, and Cybersecurity.
Another new AI-related major awaiting state approval is a collaboration between CEHC and the School of Art and Art History and the School of Music and Theater.
Source: Evangelist Jordan Carleo, State University of New York at Albany
