Collaboration with NVIDIA’s AI Infrastructure Roadmap to support joint development of next-generation memory and expand supply to accelerate global AI factory construction
News summary:
- NVIDIA and SK Hynix announced a multi-year technology partnership for next-generation memory aligned with NVIDIA’s AI infrastructure roadmap.
- This agreement will support the supply of advanced memory to address extended development cycles, advanced manufacturing, and capital investments to sustain the global build-out of AI factories.
- Diversifying into new markets that NVIDIA is creating across AI infrastructure, personal AI, and physical AI, SK Hynix will co-develop memory for NVIDIA Vera Rubin AI supercomputers, Vera CPUs, RTX Spark-powered PCs, and Jetson Thor robotic computing platforms.
- The companies will apply AI to semiconductor chip design and manufacturing using the NVIDIA CUDA-X library and NVIDIA PhysicsNeMo to accelerate semiconductor simulation, TCAD workflows, and internal engineering code.
- SK Hynix combines NVIDIA Omniverse, OpenUSD Scene Optimization, and NVIDIA cuOpt to power the factory digital twin and drive fully autonomous factory operations.
NVIDIA and SK Hynix today announced a multi-year technology partnership to advance next-generation memory and accelerate semiconductor design and manufacturing to build global AI factories. This agreement builds on years of in-depth joint engineering collaboration that has powered some of the world’s most advanced AI computing platforms.
“AI factories are the powerhouse of the next industrial revolution, and advanced memory is critical to their performance,” said Jensen Huang, founder and CEO of NVIDIA. “SK Hynix has been an exceptional partner for NVIDIA and has played a central role in delivering advanced memory technologies for NVIDIA AI computing platforms. Together, we will co-develop the next generation of memory for AI factories and support the accelerated global expansion of AI infrastructure, from frontier model training to agentic and physical AI.”
SK Group Chairman Choi Tae-won said, “SK Hynix and NVIDIA have been building toward this for years, and this partnership reflects the depth of that collaboration.” “Together, we are co-developing next-generation memory for AI factories and applying AI to how semiconductors are designed and manufactured. This is work that will shape the future of AI infrastructure.”
The multi-year agreement will support supplies to meet extended development cycles for advanced memory. As AI factories expand globally, this strategic partnership will ensure memory supplies keep pace with NVIDIA’s infrastructure roadmap and enable us to continue building AI infrastructure around the world. Through this partnership, SK Hynix will diversify into new markets created by NVIDIA (spanning AI infrastructure, personal AI, and physical AI) and co-develop memory for NVIDIA Vera Rubin AI supercomputers, NVIDIA Vera CPUs, NVIDIA RTX Spark™-powered PCs, and NVIDIA Jetson Thor™ robotic computing platforms.
Acceleration Technology Computer Aided Design and Semiconductor Simulation
SK Hynix uses NVIDIA CUDA-X™ libraries and AI to accelerate semiconductor simulations, including computer-aided design and computational lithography workflows.
SK Hynix also uses CUDA-X and the NVIDIA PhysicsNeMo™ framework to accelerate core workloads across its in-house simulation code and AI physics workflow.
By extending these tools to the semiconductor electronic design automation and simulation ecosystem, this initiative paves the way for a three-way collaboration between chipmakers, NVIDIA, and electronic design automation software vendors.
Advances in fab digital twins for automated manufacturing
SK Hynix is developing a fab digital twin as the basis for autonomous fab operations. Using scene optimization technology, the NVIDIA Omniverse™ library, and the OpenUSD pipeline, teams can build 3D factory scenes to visualize, simulate, and optimize complex semiconductor manufacturing environments.
These digital twins can also support operational optimization, including the movement of autonomous mobile robots and other fab assets, using the open source GPU-accelerated NVIDIA cuOpt™ decision optimization engine and the NVIDIA Metropolis platform.
The companies are also exploring ways to connect digital twins with existing legacy software and agent-based AI workflows, allowing AI systems to reason with fab data, automate tasks, and improve manufacturing decisions.
