LG Meets with Nvidia Executives to Develop Specialized AI Models for Specific Fields

[Financial News] LG AI Research said on the 22nd that it met with Nvidia executives on the 21st at its headquarters in Magok District, Seoul, and agreed to strengthen their technology alliance to expand the K-EXAONE ecosystem.
At the meeting, LG AI Research executives, including co-head Lim Woo-hyung and head of EXAONE Lab Jin-sik Lee, discussed strategies with Nvidia executives, including Bryan Catanzaro, vice president of applied research who visited Korea for the Nemotron developer competition, and NVIDIA Korea CEO Jeong So-young, on cooperation in developing next-generation artificial intelligence (AI) models and jointly building an AI ecosystem.
The two companies agreed to broaden their cooperation by jointly developing specialized models that combine LG's AI model EXAONE with Nvidia's Nemotron open ecosystem.
LG AI Research and Nvidia have maintained close technical cooperation throughout the development process, from EXAONE 3.0 to K-EXAONE and the multimodal AI EXAONE 4.5, which was unveiled earlier this month.
During the development of EXAONE, LG AI Research used Nvidia's Nemotron Open Dataset to secure high-quality data training, while Nvidia supported the optimization of AI model training and improved inference performance and efficiency by providing its latest graphics processing unit (GPU) Blackwell architecture, AI development platform, and software designed to enhance inference performance.
The cooperation between LG AI Research and Nvidia has already produced tangible results.
In Stanford Institute for Human-Centered Artificial Intelligence (HAI)'s recently published AI Index Report, Korea ranked third, behind the United States and China, with five notable AI models. Four of them were LG AI Research's EXAONE series: the country's first reasoning model, EXAONE Deep; the medical AI model EXAONE Path 2.0; the country's first hybrid model, EXAONE 4.0; and the national AI model K-EXAONE.
Bryan Catanzaro, vice president at Nvidia, said, "As a key partner of LG AI Research, Nvidia has worked together to make EXAONE Korea's best AI model." He added, "By combining LG's EXAONE with Nvidia's Nemotron, we will lead Sovereign AI and contribute to ecosystem expansion."
Lim Woo-hyung said, "Nvidia is a core technology partner that has worked with us on EXAONE development," adding, "We will take the cooperation between LG and Nvidia a step further by expanding it into a research and development ecosystem and create Sovereign AI results that can be felt in industrial settings."
Meanwhile, Koo Kwang-mo, CEO of LG, has been actively seeking to expand cooperation with big tech, meeting earlier this month with Alex Karp, CEO of Palantir Technologies, and with Deepak Pathak and Abhinav Gupta, co-founders of Skild AI and globally recognized experts in physical AI, one after another.
mkchang@fnnews.com Jang Min-kwon Reporter