NVIDIA CEO Jensen Huang's upcoming visit to South Korea is a pivotal moment, highlighting the country's critical role in both securing the AI chip supply chain and pioneering new AI applications.
The primary driver for this high-level visit is the persistent bottleneck in HBM (High Bandwidth Memory). HBM is a specialized, high-performance memory essential for powering advanced AI GPUs, and Korean companies SK hynix and Samsung Electronics are the world's leading producers. For months, demand for HBM has far outstripped supply, creating a major challenge for NVIDIA's ambitious product roadmap for 2026 and beyond. SK hynix has been the dominant supplier, but this concentration creates risk. Therefore, Huang's visit is a strategic move to secure future supply by deepening ties with SK hynix and, crucially, helping Samsung accelerate its qualification process for the latest HBM stacks. It's about ensuring the production lines for NVIDIA's next-generation GPUs don't stall for lack of memory.
Beyond just chips, the visit signals a major push into what's being called "physical AI." This concept involves moving AI out of data centers and into the real world—powering robots, autonomous vehicles, and smart factories. Here, LG Electronics emerges as a key potential partner. LG has been actively developing its own AI models like EXAONE and has confirmed it's in talks with NVIDIA to collaborate on robotics and mobility. For NVIDIA, partnering with a manufacturing giant like LG offers a direct path to commercialize its AI platforms in tangible, real-world products.
The timing of this visit is no coincidence and is the result of several converging factors. First, the ongoing HBM shortage, which memory makers warn could last into 2027, has made securing supply an urgent priority. Second, LG has recently demonstrated significant progress with its AI research and has openly discussed collaborations, making it a ripe partner for NVIDIA's physical AI ambitions. Third, while Samsung has lagged behind SK hynix, its recent progress in passing quality tests for HBM3E chips makes it a viable candidate for diversifying NVIDIA's supply chain, a conversation best had at the highest levels.
Investors have already taken notice, with the stock prices of SK hynix and LG Electronics soaring over the past month, far outpacing NVIDIA's. This market reaction suggests that the real value right now is perceived to be in the hardware supply chain—the memory and components—rather than just the GPU designer itself. It's a clear signal that the market anticipates significant deals to come out of this visit. In essence, Huang's trip is a masterclass in strategic management: shoring up a vulnerable supply chain while simultaneously unlocking a new frontier for growth.
- Glossary Terms
- HBM (High Bandwidth Memory): A type of high-performance RAM used in conjunction with high-end GPUs and network accelerators. It's crucial for AI workloads that require processing massive amounts of data very quickly.
- Physical AI: The application of artificial intelligence to control physical systems in the real world, such as robots, drones, and autonomous vehicles, enabling them to perceive, reason, and interact with their environment.
- P/B (Price-to-Book Ratio): A financial ratio used to compare a company's current market price to its book value. A high P/B ratio can suggest that investors expect high future growth.