- NVIDIA and SK Hynix will jointly develop advanced memory technologies for future AI computing platforms.
- The partnership expands AI-driven semiconductor design, manufacturing, and autonomous fab operations.
NVIDIA and SK Hynix, a member of the SK Group, have entered into a multiyear strategic partnership focused on developing next-generation memory technologies aligned with NVIDIA’s future AI infrastructure roadmap. The collaboration is designed to support advanced memory solutions, extended development programs, sophisticated semiconductor fabrication processes, and capital investments needed for the global expansion of AI factories. The initiative reflects the growing demand for high-performance computing infrastructure as artificial intelligence applications continue to scale across industries.
As part of the agreement, SK Hynix will expand into emerging markets being created by NVIDIA across AI infrastructure, personal AI, and physical AI ecosystems. The companies will jointly develop memory technologies for NVIDIA’s upcoming Vera Rubin AI supercomputers, Vera CPUs, RTX Spark-powered personal computers, and Jetson Thor robotic computing platforms. Through this collaboration, both organizations aim to strengthen the performance, efficiency, and scalability of future AI computing systems while supporting increasingly complex workloads.
The partnership also extends into semiconductor research, design, and manufacturing processes through the application of artificial intelligence technologies. SK Hynix and NVIDIA will leverage NVIDIA CUDA-X libraries and the NVIDIA PhysicsNeMo framework to enhance semiconductor simulations, technology computer-aided design (TCAD) workflows, and proprietary engineering software. By integrating AI-driven methodologies into chip development, the companies seek to improve modeling accuracy, accelerate innovation cycles, and optimize engineering productivity.
In manufacturing operations, SK Hynix plans to advance digital twin capabilities within its fabrication facilities by combining NVIDIA Omniverse technologies, OpenUSD scene optimization, and NVIDIA cuOpt solutions. These technologies will support the creation of highly detailed 3D factory environments that enable visualization, simulation, and optimization of semiconductor production processes. The objective is to enhance operational efficiency while moving toward increasingly autonomous semiconductor manufacturing facilities.
Factory digital twins developed through the collaboration will allow engineering teams to model complex manufacturing environments in real time. Using scene optimization technologies, OpenUSD pipelines, and Omniverse libraries, teams can evaluate production workflows, identify bottlenecks, and improve facility performance before implementing changes in physical operations. This approach is expected to support more intelligent decision-making and greater manufacturing flexibility across semiconductor fabs.
SK Hynix is already utilizing CUDA-X libraries and the NVIDIA PhysicsNeMo framework to accelerate critical simulation workloads and AI-driven physics applications within its internal engineering environment. The expanded partnership builds on these existing deployments and further integrates advanced AI technologies into semiconductor development and manufacturing activities. Together, the companies aim to drive innovation across memory technologies, AI infrastructure, and intelligent production systems as global demand for AI computing continues to increase.
Frequently Asked Questions
What is the main objective of the NVIDIA and SK Hynix partnership?
The partnership focuses on developing next-generation memory technologies and AI-enabled semiconductor manufacturing capabilities to support future AI computing platforms. Through joint development efforts, the companies will create advanced memory solutions for AI infrastructure, personal AI devices, and robotics platforms. They will also integrate artificial intelligence into semiconductor design, simulation, and fabrication processes while expanding the use of digital twins and autonomous manufacturing technologies to improve efficiency, scalability, and innovation across semiconductor production operations.
How will AI be used in semiconductor manufacturing under this collaboration?
Artificial intelligence will be applied to semiconductor chip design, simulation, engineering workflows, and factory operations to improve performance and efficiency. NVIDIA’s CUDA-X libraries, PhysicsNeMo framework, Omniverse platform, OpenUSD technologies, and cuOpt solutions will help accelerate simulations, optimize manufacturing processes, and support digital twin development. These capabilities enable semiconductor facilities to visualize complex production environments, automate decision-making, identify operational improvements, and move toward more autonomous fabrication systems while maintaining high levels of manufacturing precision and productivity.
Click above to visit the official source.