2/24/2025 Bruce Adams
CS professor Minjia Zhang has received an NSF Career Award to support his research on building highly efficient, scalable, and easy-to-use framework for training the next-generation AI4Science models on modern hardware with massive parallelism.
Written by Bruce Adams
"My central vision for the next decade is to generalize AI system technologies to broadly address major system pain points and promote the progress for large-scale scientific discoveries.”
So says CS professor Minjia Zhang, who has received an NSF Career Award for his proposal "CAREER: A Framework to Power Next-Generation AI-based Scientific Models on Hardware with Massive Parallelism." The prestigious NSF career award recognizes and supports early-career faculty. Zhang has been granted $760,000 to support his research on building a highly efficient, scalable, and easy-to-use framework for training the next-generation AI4Science models on modern hardware with massive parallelism. His project contributes to the scientific computing community by developing advanced system technologies to train science foundation models at a large scale and enabling new capabilities for science-informed models to capture complex interactions within vast scientific data. “I’m particularly excited and honored when receiving the news about this award," Zhang says.
"While many scientists are rushing to build more intelligent AI4Science models, they often struggle to fully utilize the computing power offered by modern hardware with massive parallelism. This is due to the complexity of science-informed model architectures and the challenge of involving non-CS scientists,” he explains. “To accelerate AI4Science efforts, we must make these technologies more efficient and easy to use. We must also go beyond efficiency and optimize across AI4Science models by examining the model development cycle, spanning data, algorithms, and system hardware."
"I envision a future where AI helps us understand the functions and interactions of living systems, discovers new drugs to cure diseases, fosters international collaborations, and achieves super-intelligence that enhances our ability to model and predict natural phenomena, revolutionizing the natural sciences and benefiting all of humanity,” Zhang asserts. “The NSF CAREER Award provides a foundation towards achieving this long-term vision”.
Before joining the University of Illinois Urbana-Champaign, Zhang spent seven years as a Principal Researcher and technical lead at Microsoft Research Redmond and the Microsoft AI and Research division. Zhang has established collaborations with Argonne National Lab, Oak Ridge National Lab, and Microsoft Research, which own mission-critical AI4Science models, and the proposed work will be evaluated with collaborators. The engagement with these institutions ensures the practicality of the research and supports key US science missions.
“I am fortunate to work with exceptionally talented students who are highly motivated and skilled in systems and AI. They make research enjoyable, exciting, and rewarding. I am also grateful to my amazing colleagues at The Illinois Grainger College of Engineering Siebel School of Computer and Data Science, including Nancy Amato, Lawrence Rauchwerger, Tandy Warnow, Indranil Gupta, Darko Marinov, and Tianyin Xu, for their invaluable feedback on my NSF CAREER proposal. I also highly appreciate the support from the NSF division of CNS,” Zhang says.
In addition to research advancements, the award supports Zhang’s education and participation plans for the future of a well-trained workforce in scientific discovery. He is developing new undergraduate and graduate courses at the Siebel School of Computer and Data Science that leverage advanced system technologies to accelerate scientific research.
Grainger Engineering Affiliations
Minjiaz Zhang is an Illinois Grainger Engineering professor of computer science.