Jensen Huang's Dramatic "Rescue" of the Korean Stock Market

Author: Su Yang, Tencent Technology

On June 5th, the Korean stock market experienced a "Black Friday," with the KOSPI index closing down by 5.54%. After the market opened on June 8th, the intraday decline once widened to over 8%, triggering the circuit breaker mechanism, with Samsung and SK Hynix both falling nearly 10%.

Amidst market fears, Jensen Huang's visit dramatically played the role of "market rescue."

Earlier, on the evening of June 7th, Korea local time, Jensen Huang held a "dinner gathering" with SK Group Chairman Chey Tae-won and SK Hynix CEO Gook Ro Joon.

Afterward, Jensen Huang confirmed to the onsite media: NVIDIA's newly launched Vera CPU will use SK Hynix DRAM; both sides are preparing for a "super-large-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 announced a multi-year technical cooperation agreement involving AI supercomputers extending to robotics, digital twins, and semiconductor manufacturing.

Jensen Huang even directly boasted at the press conference, "If you are an AI company shareholder, you should be happy; their prices are very low right now."

Locking in SK Hynix Memory

Vera is NVIDIA’s first dedicated data center CPU, facing competitors including Intel’s Xeon product line, AMD’s Epyc chips, and in-house projects from large cloud providers like Amazon Graviton.

In this new battlefield, NVIDIA has anchored its memory supply to SK Hynix from the very beginning.

On June 7th, NVIDIA and SK Hynix officially announced the establishment of a multi-year technical partnership to jointly develop the next-generation memory that matches NVIDIA’s AI infrastructure roadmap.

It is understood that the cooperation covers a series of products including NVIDIA Vera Rubin AI supercomputers, Vera CPUs, PCs equipped with RTX Spark, and Jetson Thor robotic computing platforms, spanning both personal and cloud-side product lines.

The announcement states that the partnership aims to ensure the supply of advanced memory to meet the long development cycles, complex manufacturing processes, and high capital investments of such products, thereby supporting the ongoing construction of global AI factories.

The announcement also listed SK Hynix’s plans to expand into several new markets that NVIDIA is pioneering, including AI infrastructure, personal AI, and physical AI.

AI Boosts Chip Manufacturing

In addition to supplying memory, SK Hynix has begun applying NVIDIA’s AI technology to its own chip design and manufacturing processes.

Similar collaborations have 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 simulations, covering areas such as technical computer-aided design (TCAD) and computational lithography.

Both sides are also pushing to extend these tools into the semiconductor electronic design automation (EDA) and simulation ecosystems, paving the way for tripartite cooperation among chip manufacturers, NVIDIA, and EDA software providers.

This means that their collaboration is no longer limited to SK Hynix’s internal use but is exploring a model that can be promoted across the entire semiconductor industry.

In manufacturing, SK Hynix is advancing the development of digital twin functions for wafer fabs, aiming for fully autonomous factory operations. This work is based on NVIDIA’s Omniverse platform. Using Omniverse libraries and OpenUSD workflows, SK Hynix can build 3D factory scenarios for visualization, simulation, and optimization of complex semiconductor manufacturing environments.

At the factory operation level, these digital twin functions can also connect to NVIDIA’s cuOpt decision optimization engine and Metropolis platform, used to schedule autonomous mobile robots and other assets within the wafer fab.

The announcement also revealed that both companies are exploring integrating digital twins with existing traditional software and intelligent AI workflows, enabling AI systems to reason based on wafer fab data, automatically execute tasks, and improve manufacturing decisions.

Laying the Groundwork Six Months in Advance

In October 2025, NVIDIA and SK Hynix announced a large-scale infrastructure cooperation.

At that time, SK Group was building an AI factory equipped with over 50k NVIDIA GPUs, with the first phase scheduled for completion by the end of 2027. Once completed, it is expected to become one of the largest AI factories in Korea.

The factory adopts a "GPU-as-a-Service" model, open to SK Group’s subsidiaries and external organizations, aiming to accelerate Korea’s industrial digital transformation and innovation.

SK Telecom is also involved in the deployment.

As NVIDIA’s cloud partner, SK Telecom plans to build an industrial AI cloud in Asia using NVIDIA RTX PRO 6000 Blackwell server-grade GPUs, with an initial deployment of over 2,000 GPUs, specifically for running Omniverse workloads, providing computing power for SK Hynix’s semiconductor manufacturing, wafer fab digital twins, and internal AI agents.

During his visit to Korea, Jensen Huang also revealed that he is in discussions with telecom companies, as future AI will rely on telecommunications networks. This aligns with SK Telecom’s participation in the collaboration.

Three Companies Share HBM4 Orders

Although NVIDIA and SK Hynix signed a multi-year technical cooperation agreement, NVIDIA has not put all its eggs in one basket regarding HBM4 supply.

Jensen Huang made it clear to reporters upon arriving in Seoul: "All three suppliers have been qualified. All three are in production, and they are competing to support Vera Rubin."

The three suppliers correspond to Samsung Electronics, SK Hynix, and Micron Technology.

At the Taipei Computer Show keynote, Huang confirmed that Vera Rubin is fully in production, with delivery planned for the third quarter of this year. This system is built around NVIDIA Vera CPUs and Rubin graphics core clusters, with each server rack system equipped with terabyte-level HBM4 memory.

Based on the actual progress of HBM4, SK Hynix remains in the lead.

Reuters reported last September that SK Hynix had already completed internal certification of HBM4 chips and established a production system for customers, aiming to complete mass production readiness of 12-layer HBM4 products by the second half of 2025. Senior analyst Kim Sunwoo of Meritz Securities predicted that, benefiting from early supply to key customers and the resulting first-mover advantage, SK Hynix’s market share in HBM would remain above 60% in 2026.

Chip Shortage to Persist for Several Years

The situation of three companies competing for HBM4 supply does not mean the supply pressure will ease.

After the dinner, Jensen Huang also gave a somewhat pessimistic outlook. He told onsite media that the shortage of storage chips will not end soon: "The entire industry supply chain—from wafers to packaging to silicon photonics—is in shortage because demand is so high. This situation will last for several years."

This statement’s background is the nearly endless consumption of advanced memory in global AI factory construction.

The shortage Jensen Huang refers to is not just a shortage of a single component but tight supply across almost every link in the industry chain. NVIDIA’s launches of Vera Rubin, promotion of AI factories, and entry into personal and physical AI domains are all increasing demand for memory. That’s why he said all three HBM4 suppliers are competing to support Vera Rubin.

No one wants to fall behind in a supply-demand crunch.

During this trip to Korea, SK Group was a key focus, but it was not the only agenda Jensen Huang had. He revealed that he had scheduled meetings with Hyundai Motor, LG, SK, Samsung, and Naver. He also disclosed that NVIDIA is actively recruiting for its new R&D center in Korea. From these signals, it’s clear that NVIDIA is systematically deepening its ties with Korea’s tech industry, with SK Group being a key but not the only part.