ST. LOUIS — The Taylor Geospatial Institute, in partnership with the University of Illinois Urbana-Champaign, is unveiling a highly anticipated, AI-powered supercomputing capability that is available to researchers at any of the eight partner institutions that make up the TGI consortium.
TGI Researchers can now access TGI RAILS to tap into top-of-the-line data analysis and storage to accelerate their geospatial science.
Made possible through a $1 million grant from the National Science Foundation to purchase hardware, the TGI Regional AI Learning System (TGI RAILS) has a petabyte of useable storage, 750 teraflops of computing power, a networking speed of 400 gigabits per second, and the ability to support the training and application of AI algorithms and machine learning models used in geospatial research. Housed at the National Petascale Computing Facility (NPCF) at Illinois, TGI RAILS—which incorporates the most advanced GPUs available, Nvidia’s H100s—has the potential to accelerate geospatial science and discovery by orders of magnitude beyond what is possible using standard computing resources used in labs across the TGI consortium.
“We are providing a new way of servicing the geospatial community. The specialized focus of this system on geospatial and AI-related geospatial work is unique,” said William Kramer, Executive Director of the New Frontiers Initiative at Illinois. “As leaders in high-performance computing, this is a great opportunity for Illinois to collaborate with the TGI consortium and organize a focus around helping people make use of this technology.”
Nadine Alameh, Executive Director of TGI, agreed. “This computing power can be transformative for the TGI community,” she said. “Especially as we focus on Geo AI/ML as a strategic priority, TGI RAILS will allow researchers across the TGI consortium to access cutting-edge, high-performance computing to analyze big datasets in unprecedented ways. I expect many novel geospatial insights to emerge from TGI RAILS, and I can’t wait to see what it can do!”
“One unique aspect of [TGI RAILS], versus other alternatives, is that it’s using cutting edge—first of its kind available when they were ordered—GPU processing units,” said Jeremy James Enos, System Management & Development Lead at the National Center for Supercomputing Applications (NCSA), which administers the NPCF.
As impressive as the TGI RAILS hardware is, so is the support model that will funnel users to the system, added Kramer. “It is an interesting experiment that we have not done before at NCSA in that the support will be distributed across the TGI institutions,” he said. “It is going to be a very integrated way of helping people make use of this system, some with local support people and, at NCSA, we’ll provide some backup for those people.”
Kramer noted that with all its unique qualities, TGI RAILS has the potential to speed the processing and analysis of geospatial data that could save researchers months of time. “Certainly, the computing power and the storage capabilities would propel discovery faster than it would without it,” he said. “This model of integrating expertise across all eight of the TGI institutions allows for more a customized type of support.”
About Taylor Geospatial Institute
TGI is passionate about sparking a revolution in geospatial science and technology to create the next generation of solutions and policies that the whole world will depend on for sustainability and growth.
The TGI consortium includes Saint Louis University, the Donald Danforth Plant Science Center, Harris-Stowe State University, University of Illinois Urbana-Champaign, Missouri University of Science & Technology, University of Missouri-Columbia, University of Missouri-St. Louis, and Washington University in St. Louis. Collectively, these institutions cover geospatial research from ocean depths to outer space.
For more information, visit taylorgeospatial.org.
MEDIA CONTACT
info@taylorgeospatial.org