Nvidia has signed a trio of technology partnerships with South Korean tech companies SK Hynix (KRX:000660), SK Telecom (KRX:017670) and Naver (KRX:035420), according to press releases from the US chip giant on Sunday.
The partnerships position South Korea as a major hub for Nvidia's AI infrastructure.
Under a new multiyear hardware agreement, Nvidia and SK Hynix will co-develop advanced memory for Nvidia's AI supercomputers, including its upcoming Vera Rubin platform.
The new Vera CPU, which handles central processing workloads for the Rubin architecture, is currently in full production. Nvidia CEO Jensen Huang told Bloomberg News that systems built on the new architecture are scheduled to begin shipping in the third quarter.
The collaboration will also focus on producing custom memory for Nvidia's Jetson Thor robotic computing platforms.
As part of the deal, the South Korean chipmaker, a rival to Samsung Electronics (KRX:005930) and Micron Technology, gains access to Nvidia software suites like PhysicsNeMo and Omniverse. SK Hynix plans to integrate these tools into its in-house simulation codes and 3D visualizations to build "digital twins" of its manufacturing facilities, optimizing autonomous fab operations and the movement of mobile robots.
Nvidia has also partnered with SK Telecom to deploy the full-stack Nvidia DSX platform.
The agreement aims to build a gigawatt-scale AI Cloud in South Korea, with the first AI data center scheduled to go online by 2027. This infrastructure will provide GPU-based cloud computing tailored to support sovereign, physical, and agentic AI workloads for enterprises across Korea, with plans to expand to Asia.
"Telecom networks are becoming national AI infrastructure," said Huang. "They connect people, companies, devices and machines - and now they can become the backbone of new AI clouds."
Separately, Nvidia also teamed up with Naver to expand sovereign AI infrastructure, starting at 55 megawatts with plans to move to gigawatt scale using the DSX platform.
Naver will leverage Nvidia's open model Nemotron 3 Ultra to further scale its HyperCLOVA X model for enterprise use.
Naver will also deploy the Nvidia Cosmos platform to accelerate its "Seoul World Model," which uses real-world Korean spatial data to train autonomous driving systems.
While the exact financial terms of these collaborations were not disclosed, Reuters reported that the hardware partnership with SK Hynix will run for at least two years, with an option to extend, following high-profile meetings between Jensen Huang, SK Group Chairman Chey Tae-won, and Naver founder Lee Hae-jin.
In Seoul trading at midday on Monday, SK Hynix's shares fell nearly 2%, Naver jumped almost 14%, while SK Telecom rose nearly 7%.



