Nvidia, Co-Founder Give $50M To University Of Florida For AI Data Center
‘They should be able to make marked advancements in science, but the use of AI, to infuse it into the workforce and to help businesses upskill and become more competitive, it‘s something that [the University of Florida] is quite interested in, and they have the resources and the reach to make that happen,’ Nvidia co-founder Chris Malachowsky says.
Nvidia and co-founder Chris Malachowsky are donating $50 million in cash, technology and services to the University of Florida for an artificial intelligence-centric data center that will house what they say will become the “world‘s fastest AI supercomputer in higher education.”
The new partnership, announced Tuesday, will also be supported by an additional $20 million investment from the university, which said the combined investment will upgrade its existing HiPerGator supercomputer with an Nvidia DGX SuperPOD that consists of 140 DGX A100 systems and 1,120 A100 GPUs, making it capable of delivering 700 petaflops of AI performance.
[Related: These Are The Top 10 Nvidia Partners Driving AI Adoption]
While Malachosky‘s $25 million donation is a financial contribution, Nvidia’s $25 million gift consists of discounts on hardware, software, services and training. As part of the partnership, Nvidia’s Deep Learning Institute will work with the university on developing new curricula and courses that will include programming for young adults and teenagers, and the university plans to hire 100 new faculty members to support the expanded class offerings.
The university will also become an Nvidia AI Technology Center, a designation that means university graduate fellows and Nvidia employees will collaborate on new advancements in AI. In addition, Nvidia solution architects and engineers will work with the university on the installation, operation and optimization of Nvidia-based high-performance computing resources.
Malachowsky, who graduated from the University of Florida in 1983 and met his future wife there, said the expanded HiPerGator system will likely be among the 10 fastest AI-focused supercomputers in the world when it goes online in early 2021. He said the university plans to make the system widely available to students and researchers throughout the state of Florida, particularly universities with fewer resources, to work on “state-of-the-art equipment and state-of-the-art problems.
“They should be able to make marked advancements in science, but the use of AI, to infuse it into the workforce and to help businesses upskill and become more competitive, it‘s something that [the University of Florida] is quite interested in, and they have the resources and the reach to make that happen,” Malachowsky said.
Joe Glover, University of Florida‘s provost, said many faculty members are already involved in AI research and development, from autonomous vehicles to drug discovery. One of the projects that will benefit from the expanded HiPerGator system is the university’s iCoast initiative, which is studying the rising sea levels along Florida’s coastline and generating simultaneous data streams on a variety of things, including water temperature, salinity, chemistry and biological life.
“We launched this iCoast initiative before this remarkable AI opportunity came along, and we are in the process of generating massive amounts of data that absolutely require the capabilities of this machine to completely understand,” he said.
Dominic Daninger, vice president of engineering at Nor-Tech, a Burnsville, Minn.-based Nvidia HPC partner, said HPC has become increasingly important for numerous fields of research, most notably for potential treatments and therapies for the novel coronavirus in recent months, and Nvidia‘s data center GPUs are becoming more important for accelerating such work.
“We‘re seeing that it’s very important,” he said. “Some of the early work done on COVID certainly sped up the process of finding potential therapies and vaccine formulations,” he said.