Despite facing cuts to academic research under the Trump administration, the US National Science Foundation (NSF) is financing a new $20-million supercomputer built by the Georgia Institute of Technology that will use AI to advance science, the university said Tuesday.
The supercomputer, called Nexus, is intended to find new cures for diseases, better understand the human brain, and advance quantum computing. It will calculate more than 400 quadrillion operations per second. Georgia Tech will retain 10% of its capacity, while researchers from around the US can apply to use it as well.
The funding is one of the more significant commitments by the NSF since the Trump administration proposed slicing its budget by 57% to $3.9 billion, significantly limiting the number of grants it can award and the impact federal funds could have on academic AI research. The administration also paused new funding, laid off staff involved in AI research, and canceled hundreds of existing grants. Some of the cuts are tied up in court, making the NSF’s funding status uncertain.
With the likelihood of a constrained budget, the NSF’s decisions on project funding carry much greater weight. Investing in a supercomputer can advance innovations across science, math, technology, and other disparate disciplines, where projects with much more focused audiences may see fewer dollars coming their way. The initiative also reinforces the critical role of university research in AI innovation, which has been undermined by the cuts.