The South Korean equity market faced severe volatility on June 5, marking a 'Black Friday' where the KOSPI index plummeted 5.54%. By June 8, intraday losses expanded beyond 8%, triggering trading circuit breakers and dragging Samsung and SK Hynix down nearly 10%. Amidst this turmoil, NVIDIA CEO Jensen Huang's intervention stabilized sentiment, effectively acting as a market savior. On the evening of June 7, Huang convened a strategic dinner with SK Group Chairman Chey Tae-won and SK Hynix CEO Lee Seok-Hee. Following the meeting, Huang confirmed to assembled media that NVIDIA's upcoming Vera CPU will utilize SK Hynix DRAM, signaling a large-scale cooperation extending through the second half of the current year and into the next. He further cautioned that the prevailing memory chip shortage would endure for several years.

NVIDIA and SK Hynix subsequently formalized a long-term technology partnership encompassing AI supercomputing, robotics, digital twins, and semiconductor manufacturing. During a press conference, Huang explicitly buoyed investor confidence, stating, 'If you are a shareholder of an AI company, you will be happy because their prices are currently very low.' The Vera CPU represents NVIDIA's first standalone data center processor, designed to compete directly with Intel's Xeon line, AMD's Epyc chips, and custom silicon from cloud giants like Amazon's Graviton. In this emerging competitive landscape, NVIDIA has anchored its memory supply chain to SK Hynix from the outset. The June 7 announcement detailed a joint development roadmap for next-generation memory aligned with NVIDIA's AI infrastructure, covering product lines from the Vera Rubin AI supercomputer and Vera CPU to the RTX Spark PC and Jetson Thor robot computing platform.

The collaboration aims to secure advanced memory supplies against the backdrop of extended development cycles, complex manufacturing processes, and high capital expenditures required for global AI factory construction. SK Hynix is set to expand into new markets led by NVIDIA, including AI infrastructure, personal AI, and physical AI. Beyond mere supply, SK Hynix is integrating NVIDIA's AI technology into its own chip design and manufacturing workflows. This mirrors a prior collaboration with TSMC focused on computational lithography. Data compiled by Woofun AI indicates that SK Hynix is leveraging NVIDIA's CUDA-X library and AI to accelerate semiconductor simulations, specifically targeting technology computer-aided design (TCAD) and computational lithography. The partnership extends these tools to the semiconductor electronic design automation (EDA) and simulation ecosystem, paving the way for a trilateral model involving chip manufacturers, NVIDIA, and EDA software providers.

In the manufacturing domain, SK Hynix is advancing wafer-level digital twinning capabilities to achieve fully autonomous factory operations, built upon the NVIDIA Omniverse platform. Utilizing the Omniverse kit and OpenUSD workflow, the company can construct 3D factory scenarios to visualize, simulate, and optimize complex semiconductor environments. Operationally, these digital twins integrate NVIDIA's cuOpt decision optimization engine and Metropolis platform to schedule autonomous mobile robots and assets within the wafer fab. The companies are also exploring the integration of digital twinning with legacy software and intelligent agent AI workflows, enabling AI systems to infer, execute tasks autonomously, and refine manufacturing decisions . Woofun AI notes that this approach transforms the collaboration from internal utility to an industry-wide scalable model.

In October 2025, NVIDIA and SK Hynix accelerated a large-scale infrastructure collaboration by six months. The SK Group is constructing an AI factory equipped with over 50,000 NVIDIA GPUs, with the first phase targeted for completion by the end of 2027. Upon completion, this facility is projected to become one of South Korea's largest AI factories, operating on a 'GPU as a service' model open to SK Group subsidiaries and external entities to drive industrial digital transformation. SK Telecom, an NVIDIA cloud partner, is participating in specific deployment efforts. The carrier plans to utilize NVIDIA RTX A6000 Blackwell server-grade GPUs to build an Industrial AI Cloud in Asia, with an initial deployment of over 2,000 GPUs dedicated to Omniverse workloads. This infrastructure will provide computational support for SK Hynix's semiconductor manufacturing, wafer fab digital twins, and internal AI agents.

During his visit, Huang revealed ongoing discussions with telecommunications firms, anticipating that future AI will rely heavily on telecom networks, a direction aligned with SK Telecom's involvement. Despite the multi-year agreement with SK Hynix, NVIDIA maintains a diversified strategy for HBM4 supply. Upon arriving in Seoul, Huang stated to reporters, 'All three suppliers have been qualified. All three are in production, and they are fighting to support Vera Rubin.' These suppliers are Samsung Electronics, SK Hynix, and Micron Technology. At Computex Taipei, Huang confirmed that the Vera Rubin system is in full production with deliveries scheduled for the third quarter of this year. Each server rack system pairs the NVIDIA Vera CPU and Rubin graphics core cluster with TB-level HBM4 memory.

Regarding HBM4 progress, SK Hynix retains a leadership position. Reuters reported in September of the previous year that SK Hynix completed internal certification of HBM4 chips and established a customer production system, aiming for mass production of 12-layer HBM4 products in the second half of 2025. Meritz Securities senior analyst Kim Sunwoo predicted at the time that SK Hynix's HBM market share in 2026 would remain just over 60% due to early supply advantages. With a three-way supplier dynamic, supply-side pressure is not expected to ease. Woofun AI analysis suggests that the shortage cited by Huang is not material-specific but a systemic tightness across the entire industry chain, from wafers to packaging and silicon photonics. The insatiable demand driven by global AI factory construction, the Vera Rubin launch, and expansion into personal and physical AI sectors ensures this scarcity will persist for years.

While the SK Group was a focal point, Huang's itinerary included meetings with Hyundai, LG, SK, Samsung, and Naver. He also disclosed active recruitment for a new NVIDIA R&D center in Korea. These movements indicate a systematic deepening of ties with the entire Korean tech industry. The SK Group remains a critical component, but NVIDIA's strategy encompasses a broader ecosystem engagement to secure the hardware foundation for the next generation of artificial intelligence.