Nvidia's move to establish joint laboratories with leading Korean universities for artificial intelligence (AI) research is strongly aimed at securing engineering talent, according to analysts. It is also seen as an effort to create a testing ground for applying physical AI technologies across various fields in Korea's manufacturing-strong industrial sector.
The research area to be pursued by Nvidia and the Korea Advanced Institute of Science and Technology (KAIST) is practical AI, according to academia on the 24th. The projects will develop industry-specialized AI models as well as large-scale AI models optimized for Korean and agentic AI models. For organic collaboration, Dr. Kim Hyun-woo, formerly of Nvidia, will join KAIST's Kim Jaechul Graduate School of AI to serve as co-director of the joint laboratory. As Kim has researched physical AI technologies that interact with humans, he is expected to focus on implementing AI systems suited to Korea's manufacturing environment.
The establishment of the joint laboratory will also expand personnel exchange. The KAIST joint lab will support at least 10 of its own researchers each year and provide Nvidia internship opportunities to the supported researchers. Nvidia plans to hire outstanding Korean AI researchers for full-time research positions. Seoul National University will also engage in exchange by having students participate in Nvidia's internship program and Nvidia researchers reside on the university's campus. Furthermore, Seoul National University and Nvidia plan to jointly develop an AI education curriculum.
Seoul National University's Robotics Research Institute, scheduled to officially launch in August, has also strengthened cooperation with Nvidia. Following the May visit of Madison Huang, Nvidia's senior director of Omniverse and Robotics and daughter of CEO Jensen Huang, Nvidia provided edge (on-site deployment) AI computing modules for robots to the university free of charge and conducted training on software development applications. Discussions are also underway to incorporate parts of Nvidia's Deep Learning Institute (DLI) program, its AI education program for professionals, into classes at Seoul National University's Department of Mechanical Engineering.
The AI industry views Nvidia as seeking to expand Korea's graphics processing unit (GPU)-centered AI development ecosystem by strengthening research with domestic universities, while at the same time identifying the human resources it needs. The 260,000 GPUs that Nvidia previously pledged to supply to Korea are expected to serve as a catalyst for AI research not only at companies and government agencies but also in academia. A Seoul National University professor who requested anonymity said, "The reputation of Korean employees at Nvidia's headquarters in San Jose, U.S., is quite good," adding, "They are trying to bring in engineering talent from Korea's top universities."
