SK hynix on the 8th officially declared a next-generation memory co-development and a strengthened long-term technology partnership with Nvidia to build global AI infrastructure.
The two companies agreed to advance the partnership based on years of collaboration on world-class AI computing platforms. The cooperation goes beyond a simple supply contract and is drawing industry attention as a technology alliance that applies AI across semiconductor design and manufacturing.
On the day, Nvidia founder and CEO Jensen Huang emphasized, "AI factories are the engine of the next industrial revolution, and advanced memory is central to their performance," and said, "SK hynix has been an outstanding partner, playing a key role in providing advanced memory technology for the Nvidia AI computing platform." Huang noted, "From frontier model training to agentic AI and physical AI, we will jointly support accelerating the expansion of global AI infrastructure."
Chey Tae-won, chairman of SK Group, said, "This partnership proves the depth of collaboration the two companies have built over years," adding, "By co-developing next-generation memory for AI factories and applying AI to semiconductor design and manufacturing, we will shape the future of AI infrastructure together."
The core of the long-term partnership is securing supply stability. Considering the long development cycle unique to advanced memory, the plan is to establish a stable supply system linked to Nvidia's infrastructure roadmap. SK hynix plans to use this to proactively respond to global demand for building AI infrastructure.
The scope of business will also expand broadly. Leveraging this agreement, SK hynix will make a full-fledged entry into new AI markets pioneered by Nvidia, including AI infrastructure, personal AI, and physical AI. The product lines the two companies said they will co-develop include ▲ Vera Rubin AI supercomputer ▲ Vera CPU ▲ RTX Spark PC ▲ memory for the Jetson Thor robotic computing platform, spanning the AI hardware ecosystem from data centers to edge and robotics.
The axis of cooperation does not end with memory supply. The two companies will also jointly advance simulation technologies essential for semiconductor development. SK hynix is using Nvidia's CUDA-X libraries and the "PhysicsNeMo" framework to improve the speed and efficiency of simulations across semiconductor design and manufacturing. The applied technologies include TCAD (technology computer-aided design) for analyzing semiconductor process characteristics and computational lithography for fine circuit implementation.
Advancing the Digital Twin, cited as a key task for realizing autonomous manufacturing, is also a major agenda item of this cooperation. SK hynix is using Nvidia's Omniverse and OpenUSD technologies to reproduce actual semiconductor fabs in three-dimensional virtual space and to precisely visualize, analyze, and optimize complex manufacturing environments.
The two companies will also work on linking this Digital Twin environment with existing manufacturing systems and AI-based work processes. The plan is to have AI analyze fab operations data in real time and automate tasks to achieve improvements in productivity and operational efficiency and to advance manufacturing decision-making.