Located in the Jacobs School of Engineering’s Fusion Engineering Institute at UC San Diego, the Fusion Data Science and Digital Engineering Center is the most recent collaboration between the San Diego Supercomputer Center and General Atomics. The FDSDEC recently hit its one-year mark.

Established in March 2025, the AI learning and computing facility focuses on creating efficient products with clean fusion energy. GA, a San Diego-based energy company and weapons manufacturer, is a long-standing partner of the SDSC and has established multiple programs at UCSD.

According to professor and SDSC Director Frank Wurthwein, the new center has no ties or involvement with GA’s weaponry department. He says the SDSC works solely with GA’s fusion sector.

This newly implemented center is aimed at strengthening existing UCSD and GA programs — such as the Fusion Innovative Research Engine, funded by the Department of Energy. One collaborative project funded by FIRE includes the Fusion Energy Data Ecosystem and Repository.

FEDER is a large digital library of sourced research and experimental simulations. It is composed of findings and workflows from labs and universities across the globe that scientists and researchers can access to expedite their research and scientific discoveries. In hopes of ultimately expanding the goal of commercial fusion with data fusion AI platforms, this program connects the scientific community, fast-tracks fusion development with shared results, and increases experimental production.

“[We’re] building a data commons for a wide range of fusion science initiatives and engineering initiatives,” Wurthwein said. “And by bringing them all under one umbrella, all of the data then becomes available for all kinds of scientific programs.”

In the past year, the FDSDEC has achieved notable achievements in industry experimentation by increasing the accessibility of high-end technology to fusion researchers.

“[It] allows a scientist anywhere in the U.S. who wants to study a specific fusion problem, or choose the engineering problem, to study this within the context of using all of this data together,” Wurthwein said. “And that’s the collaboration itself. And so, it’s making breakthroughs and fields in different industries.”

Wutherwein emphasized, however, that the past year’s real achievement stems from the data fusion center’s first digital twin prototype of the DIII-D Tokamak.

The DIII-D Tokamak is a physical, high-tech, donut-shaped reactor with a circulating magnetic field that confines plasma and creates large amounts of fusion energy, similar to the sun. It takes powerful levels of fusion energy and AI to create experimental digital simulations for advanced research in fusion and engineering.

“You put it onto a computer and simulate the entirety,” Wurthwein said. “And if you can devise fast enough simulations, you can actually use that simulation to explore what should [one] have as configurations of the physical experiment, and you can get some of the trial and error out of it.”

Researchers can use the digital twin of the DIII-D Tokamak to prepare their fusion research for the physical system. Its development opens up a wide range of research roles for students from across the University of California, California State University, community colleges, and private universities.

Since its inception, the center has also partnered with Jacobs’ Fusion Engineering Institute to develop clean energy pathways for California. By implementing FDSDEC technology in the FEI, the facility provides research and study opportunities for students and staff.

UCSD’s facility has attracted significant attention across the industry. AI chip company Nvidia and a leading energy simulation software suite, Ansys, have turned to the FDSDEC as they seek research investment.

“Given that we want to build digital twins for a fusion device, [Ansys was] very interested in this as a collaboration because they would like to learn from us how to work with us on building digital twins of fusion devices and then be in the position to commercialize it,” Wurthein said. “And given that we are trying to accelerate fusion with AI … [Nvidia is] very interested in supporting this as a future customer base.”

Nvidia’s net worth is at an estimated value of $4 trillion, and with Ansys valued at around $31 billion, contributions to the center by AI industry’s top leaders leaves developers and research staff eager for partnership with these companies.

“In the ideal world, we want to have a partnership that includes industry,” Wurthwein said. “And the first industry partners that were already there at inception, when the center was created, were in Nvidia and Ansys.”

This collaboration has brought an increase in research opportunities for students and staff alike and has created technological advancements for a cleaner future for fusion energy. SDSC aspires to explore future partnerships with GA to work with UCSD’s students and staff research departments to accelerate scientific developments on campus.

UPDATE 4/20: A quote was updated for accurate transcription; Frank Wurthwein’s title was corrected; and information about the DIII-D Tokamak digital twin was corrected, following an email request from the San Diego Supercomputer Center.