Author: Su Yang
Editor: Xu Qingyang
On June 5, the South Korean stock market experienced a "Black Friday," with the KOSPI index plunging 5.54% at the close. On June 8, after opening, the intraday decline expanded to over 8%, triggering the exchange's circuit breaker mechanism. Both Samsung and SK Hynix fell nearly 10%.
Amidst market panic, Huang Renxun's visit dramatically assumed the role of "saving the market."
Prior to this, on the evening of Sunday, June 7, local time in South Korea, Huang Renxun held a "dinner meeting" with SK Group Chairman Choi Tae-won, SK Hynix CEO Kwak Noh-jung, and others.
After the meal, Huang Renxun confirmed to on-site media: Nvidia's newly launched Vera CPU will adopt SK Hynix DRAM; both sides are preparing for "hyper-scale cooperation" in the second half of this year and next year; regarding the current memory chip shortage, he believes it will last for several years.
Subsequently, Nvidia and SK Hynix officially announced a multi-year technical cooperation agreement, involving extending AI supercomputing to robotics, digital twins, and semiconductor manufacturing.
Huang Renxun even directly hyped stocks at the press conference, stating, "If you are an AI company shareholder, you would be happy; currently, their prices are very low."
01 Locking in SK Hynix Memory
Vera is Nvidia's first independent, dedicated data center CPU. Its competitors include Intel's Xeon product line, AMD's Epyc chips, and large cloud service providers' self-developed projects like Amazon's Graviton.
On this new battlefield, Nvidia has from the outset anchored its memory supply to SK Hynix.
On June 7, Nvidia and SK Hynix officially announced the establishment of a multi-year technical partnership to jointly develop next-generation memory matched with Nvidia's AI infrastructure roadmap.
It is understood that the collaboration covers a series of personal and cloud product lines including Nvidia's Vera Rubin AI supercomputer, Vera CPU, PCs equipped with RTX Spark, and the Jetson Thor robotics computing platform.
The announcement pointed out that the cooperation aims to secure the supply of advanced memory to cope with the long development cycles, complex manufacturing processes, and high capital investment of such products, thereby supporting the continuous construction of global AI factories.
The announcement also listed several new markets that SK Hynix will expand into, which Nvidia is pioneering, including AI infrastructure, personal AI, and physical AI.
02 AI Repays Chip Manufacturing
Beyond supplying memory, SK Hynix has begun incorporating Nvidia's AI technology into its own chip design and manufacturing processes.
Similar cooperation has previously been implemented at TSMC, most notably in "computational lithography."
According to the announcement, SK Hynix is using Nvidia's CUDA-X libraries and AI to accelerate semiconductor simulation, covering areas such as Technology Computer-Aided Design (TCAD) and computational lithography.
Both parties are also working to extend these tools into the semiconductor Electronic Design Automation (EDA) and simulation ecosystem, paving the way for tripartite collaboration among chip manufacturers, Nvidia, and EDA software vendors.
This means the collaboration is no longer limited to SK Hynix's own use but is exploring a model that can be promoted to the entire semiconductor industry.
In the manufacturing segment, SK Hynix is advancing the development of digital twin functions for its wafer fabs, with the goal of achieving fully autonomous factory operations. This work is based on the Nvidia Omniverse platform. Leveraging the Omniverse library and OpenUSD workflows, SK Hynix can construct 3D factory scenes for visualizing, simulating, and optimizing complex semiconductor manufacturing environments.
At the factory operations level, these digital twin capabilities can also integrate with Nvidia's cuOpt decision optimization engine and Metropolis platform to schedule autonomous mobile robots and other assets within the fab.
The announcement also revealed that the two companies are exploring connecting digital twins with existing traditional software and agentic AI workflows, enabling AI systems to reason based on fab data, automatically execute tasks, and improve manufacturing decisions.
03 Paving the Way Half a Year in Advance
In October 2025, Nvidia and SK Hynix announced a large-scale infrastructure collaboration.
At that time, SK Group was building an AI factory equipped with over 50,000 Nvidia GPUs, with the first phase scheduled for completion by the end of 2027. Upon completion, this is expected to become one of South Korea's largest AI factories.
This factory adopts a "GPU as a Service" model, open to SK Group subsidiaries and external organizations, aiming to accelerate the digital transformation and industrial innovation of South Korean industries.
SK Telecom is also undertaking specific deployment work within this framework.
As a cloud partner of Nvidia, SK Telecom plans to build an industrial AI cloud in Asia using Nvidia RTX PRO 6000 Blackwell server-edition GPUs, with an initial deployment scale of over 2,000 GPUs, specifically running Omniverse workloads to provide computing power for SK Hynix's semiconductor manufacturing, fab digital twins, and internal AI agents.
During his visit to South Korea, Huang Renxun also revealed a piece of information: he is in discussions with telecommunications companies because future AI will utilize telecom networks. This aligns with the direction of SK Telecom's involvement in the collaboration.
04 Three Companies Split HBM4 Orders
Although Nvidia signed a multi-year technical cooperation agreement with SK Hynix, it did not put all its eggs in one basket for HBM4 supply.
Upon arriving in Seoul, Huang Renxun clearly stated to reporters: "All three suppliers have been qualified. All three suppliers have entered production, and they are all racing to support Vera Rubin."
These three suppliers correspond to Samsung Electronics, SK Hynix, and Micron Technology.
During his keynote speech at the Taipei International Computer Show, Huang Renxun confirmed that Vera Rubin is in full production and is scheduled for delivery in the third quarter of this year. This system is built around Nvidia's Vera CPU and Rubin graphics core clusters, with each server rack system paired with terabytes of HBM4 memory.
Looking at the actual progress of HBM4, SK Hynix remains in the leading position.
Reuters reported in September last year that SK Hynix had completed internal certification of HBM4 chips and established a production system for customers, aiming to complete mass production readiness for 12-layer HBM4 products in the second half of 2025. Kim Sunwoo, a senior analyst at Meritz Securities, predicted at the time that thanks to early supply to key customers and the resulting first-mover advantage, SK Hynix's HBM market share in 2026 would remain at just over 60%.
05 Chip Shortage to Last for Years
The situation of spreading HBM4 supply among three companies does not mean supply pressure has eased.
After Sunday's dinner, Huang Renxun also gave a less optimistic assessment. He told on-site media that the memory chip shortage would not end soon. "The entire industry supply chain—from wafers to packaging to silicon photonics... everything is in shortage because demand is so high. This situation will last for years."
This statement was made against the backdrop of the global construction of AI factories' seemingly endless consumption of advanced memory.
The shortage Huang Renxun mentioned is not about lacking one specific component, but almost every link in the industrial chain is under supply pressure. Nvidia's launch of Vera Rubin, promotion of AI factories, and entry into personal AI and physical AI are all driving up demand for memory. This is also why he said the three HBM4 suppliers are all racing to support Vera Rubin.
No one wants to fall behind in a situation of undersupply.
For this South Korea trip, while the SK Group was a focus, it was not Huang Renxun's entire schedule. Upon arrival, he revealed that he had arranged meetings with Hyundai Motor, LG, SK, Samsung, and Naver. He also disclosed that Nvidia is actively hiring for its new R&D center in South Korea. From these moves, it appears Nvidia is systematically deepening its ties with the entire South Korean tech industry. The SK Group is a key part of this, but not the only part.
