SK Hynix Chief Executive Kwak Noh-jung has warned that the global memory chip industry is poised to face its most severe supply shortage in 2027, with demand expected to outpace production capacity well into the next decade. Kwak made these remarks during an interview coinciding with the South Korean semiconductor company’s debut on the Nasdaq stock exchange on July 10.

Kwak emphasized that customer demand for memory chips continues to increase, while manufacturing capacity remains constrained. He forecast that demand would surpass supply capacity beyond 2030, despite ongoing efforts and significant expansions aimed at mitigating the imbalance.

SK Hynix’s shares surged 13 percent on their first day of trading, closing at $168 per American depositary share, above the initial price of $149. The offering raised $26.5 billion, marking one of the largest stock sales globally in recent years. The company has become a key player in the artificial intelligence (AI) hardware supply chain, notably through its development of high-bandwidth memory (HBM) modules used in Nvidia’s chipsets. This segment has fueled a surge in profits, with SK Hynix reporting a record operating profit of 47 trillion won ($31 billion) in 2025, double that of the previous year. Analysts anticipate even stronger results for the second quarter of 2026.

Looking ahead, Kwak indicated that SK Hynix is evaluating potential sites for expanding wafer fabrication, with the United States, Japan, and Southeast Asia all under consideration. Site selection will depend on the availability of sufficient land, electricity, water, and skilled labor, as well as competitive manufacturing costs. The company currently operates main production facilities in Icheon and Cheongju, South Korea, and is constructing a large plant in Yongin.

The South Korean government has announced plans to double the country’s memory chip production capacity within five years, involving investments of 400 trillion won ($266 billion) each from SK Hynix and Samsung Electronics. However, some investors have expressed concern that such aggressive expansion could increase exposure to economic downturns.

In the U.S., SK Hynix is investing approximately $4 billion to build an advanced chip packaging factory in Indiana and allocating $10 billion to develop an AI solutions company, aiming to identify new growth opportunities related to AI technologies.

While some industry observers speculate that the current AI investment cycle may be approaching a turning point—citing efforts by companies like Apple to diversify semiconductor suppliers to include Chinese manufacturers, and Meta’s attempts to monetize excess AI computing capacity—executives and analysts largely agree that memory supply shortages will persist. Nvidia’s CEO Jensen Huang recently underscored that AI memory shortages are expected to continue for several years, with SK Hynix remaining a primary supplier. Research firms UBS and Bank of America also project sustained undersupply of dynamic random-access memory (DRAM) at least until mid-2028.

Bank of America projects global hyperscaler capital expenditure to rise to about $851 billion this year and increase further to $1.15 trillion next year, driven by stronger cloud backlogs and increasing demand for AI workloads. The bank noted that the $244 billion raised by major hyperscalers in 2026 primarily reflects balance-sheet optimization rather than funding shortages.

Supporting this outlook, Micron Technology recently announced plans to increase its investment in U.S. semiconductor manufacturing to more than $250 billion through 2035, citing growing AI-related demand and incentives to bolster domestic production.

Overall, SK Hynix’s strategic focus on AI-related memory products and capacity expansion illustrates the industry’s broader efforts to respond to rapidly growing demand, even as supply constraints and market uncertainties persist.