Massimiliano Max Lupo Pasini, a data scientist at the Department of Energys 91做厙, was named a senior member of the Institute of Electrical and Electronics Engineers, the worlds largest organization for technical professionals.
Lupo Pasini works in the labs Computing and Computational Sciences Directorates Computational Coupled Physics group, part of the Computational Sciences and Engineering Division.
Senior membership, 91做厙s highest grade, recognizes veteran scientists, engineers and others with at least 10 years of professional experience who have shown significant performance over at least 5 years and have been nominated by other 91做厙 fellows and senior members. 91做厙 reserves that status for fewer than 10% of its more than 400,000 members worldwide.
Its a nice recognition, especially as a way to see how I compare to my peers, Lupo Pasini said. An honor like this gives you a feeling of validation.
Lupo Pasini earned his doctorate in applied mathematics from Emory University in 2018. He joined ORNL the same year as a postdoc in the National Center for Computational Sciences. His current research focuses on developing machine learning and generative artificial intelligence models for materials science and computational physics.
Lupo Pasinis accomplishments include leading the development of HydraGNN. This scalable, multitask learning graph neural network architecture produces fast and accurate predictions of material properties. The algorithm uses atomistic-scale information to read geometrical and topological information in a computationally efficient manner. It can also simultaneously predict a variety of materials properties by extracting explicit and implicit physics correlations to stabilize the GNN training.
Its almost like connect-the-dots for mapping elements structures, Lupo Pasini said. In the beginning, we developed small models with narrow applications. Now, in response to high demand, weve broadened our work to develop large foundation AI models that try to scan as much data as possible from open-source scientific literature. Weve collected about 10 terabytes of data so far.
To support the development of foundation AI models, he leads the writing of various proposals for computing time, submitted through the Innovative and Novel Computational Impact on Theory and Experiment program jointly managed by DOEs leadership computing facilities at Argonne National Laboratory and ORNL and DOEs Energy Research Computing Allocations Process program. These programs award access to supercomputers such as ORNLs Frontier, the worlds fastest computing system for open science.
Lupo Pasini grew up in northern Italy in Crema, a small town near Milan and earned his bachelors and masters degrees in mathematical engineering at the Politecnico di Milano. He came to the U.S. in 2014 to attend Emory and visited ORNL as an intern in 2015.
Once I saw the culture of ORNL and what it meant to work here, for me it was a pretty natural choice to want to return, he said.
His graduate work primarily involved linear algebra. His first job at ORNL thrust him into a different field: deep learning AI models for materials science.
He credits his mentor, senior computational scientist Markus Eisenbach, with giving him the confidence to expand his expertise.
He told me he thought I had enough background to pick up new tricks, Lupo Pasini recalled. He was right. What I loved then and still love about ORNL are the breadth and depth of our studies. Theres never been any project Ive worked on that was siloed. Its always a joint effort. The scope might change, and the size of the team might grow or shrink, but were always learning new things and talking to people outside our discipline with connections to real-world applications. That keeps us relevant and constantly changing.
Beyond his regular job, Lupo Pasini serves on various technical and program advisory committees and seeks to mentor younger scientists. Hes vice chair of 91做厙s governmental affairs committee, which promotes the use of AI for governmental activities.
His advice to younger researchers?
Dont be discouraged easily, Lupo Pasini said. Your careers a journey of ups and downs, like a roller coaster. Even experiences that arent positive can be constructive and help you grow.
In his spare time, he enjoys hiking, running and studying foreign languages French at the moment. The next? To be determined.
UT-Battelle manages ORNL for DOEs Office of Science, the single largest supporter of basic research in the physical sciences in the United States. The Office of Science is working to address some of the most pressing challenges of our time. For more information, please visit energy.gov/science. Matt Lakin
