DEEPX, an ultra-low-power artificial intelligence (AI) chip company, said Tuesday it is moving forward with a collaboration with Hyundai Motor Group's Robotics Lab to develop a physical AI computing platform for next-generation robots. The two companies plan to jointly develop a next-generation AI computing architecture capable of running large-scale generative AI models in real time inside robots, and to build a physical AI platform for robotics based on this technology.
The collaboration will leverage DEEPX's next-generation AI chip, DX-M2. The chip is slated for production on Samsung's 2-nanometer process. It is being developed as a computing platform designed to run AI in real time across physical AI environments, including robots, autonomous mobile systems, and industrial automation. In particular, its ultra-low-power, high-performance AI computing architecture, engineered to perform AI operations directly inside robots, is expected to significantly enhance robotic autonomy and response speed.
DEEPX and Hyundai Motor Group's Robotics Lab have jointly developed "edge brain" technology for robots based on low-power AI chips over the past three years. The latest collaboration expands the scope of their partnership into next-generation robot AI technology, building on those achievements. The two companies plan to continue expanding their cooperation on AI computing technology across a range of robot platforms.
"In the era of physical AI, ultra-low-power computing technology capable of running AI in real-world systems such as robots, vehicles, and industrial equipment will become core infrastructure," DEEPX CEO Lokwon Kim said. "Based on our ultra-low-power AI chip technology, DEEPX aims to become a global leader in physical AI computing platforms that run AI in robots and industrial systems."
Hyun Dong-jin, executive director and head of Hyundai Motor Group's Robotics Lab, added, "In the physical AI era, robots are becoming the closest point of contact between AI technology and people. The Robotics Lab aims to create robots that can naturally coexist alongside people — robots that are livable and usable. To achieve this, we are strategically building an ecosystem of core technologies, including on-device AI computing, with specialized partner companies at home and abroad across various fields."
