SK Hynix commits $38 billion to new memory chip plants as AI demand reshapes semiconductor industry

Reviewed byNidhi Govil

11 Sources

Share

SK Hynix approved a $38.1 billion investment to build two new fabrication plants in South Korea, targeting DRAM and NAND production for AI infrastructure. The Yongin Y2 and Cheongju M17 fabs will begin construction in 2027, with cleanrooms opening in 2028-2029 as the company races to address what its CEO calls the worst memory supply shortage in industry history.

SK Hynix Commits $38.1 Billion to Expand Memory Chip Manufacturing

SK Hynix has approved a 54 trillion Korean won investment, equivalent to $38.1 billion, to construct two massive memory chip fabrication plants in South Korea

1

. The board allocated 35.2 trillion won for the Yongin Y2 fab and 19.1 trillion won for the Cheongju M17 facility

3

. This marks one of the semiconductor industry's largest single investments as memory chip manufacturers scramble to meet explosive AI demand that's reshaping global supply chains.

Source: SiliconANGLE

Source: SiliconANGLE

The SK Hynix investment forms part of a broader 700 trillion won master plan unveiled last year, which dedicates 600 trillion won to the Yongin Semiconductor Cluster and 100 trillion won to expanding the Cheongju production base

2

. The South Korean tech giant has accelerated its timeline to complete all four planned Yongin fabs by 2033, pulling the original 2045 target forward by 12 years

3

.

Yongin Y2 Fab Targets High-Bandwidth Memory Production

The Yongin Y2 fab represents the second of four new chip fabs planned for the Yongin Semiconductor Cluster, located approximately 20 miles south of Seoul

4

. Construction will break ground in July 2027, with the first cleanroom scheduled to open in June 2029

1

. The facility spans 279 acres and will focus exclusively on DRAM and NAND production, particularly high-bandwidth memory and next-generation DRAM products critical for AI infrastructure

4

.

SK Hynix dominates the HBM market as the world's largest producer of this specialized memory format

4

. Companies like Nvidia, AMD, and other AI accelerator manufacturers are struggling to secure sufficient capacity to develop and produce new GPUs and data center hardware

2

. The company introduced its most advanced HBM device, a 12-layer HBM4 chip, last September, offering more than twice the bandwidth of current memory powering today's graphics cards

4

.

Cheongju M17 Addresses Surging NAND Flash Demand

The Cheongju M17 fab will focus on NAND flash memory production, with construction beginning in February 2027 and the first cleanroom operational by December 2028

1

. SK Hynix selected this location strategically because the site already houses three existing fabs—M11, M12, and M15—along with established power and water infrastructure that will accelerate construction timelines

3

.

Source: TechSpot

Source: TechSpot

Demand for NAND flash is surging rapidly, driven primarily by hyperscalers and enterprise SSDs being deployed to expand AI service capacity

2

. NAND chips serve as cache storage in AI inference workloads, with agentic AI and physical AI applications like robots expected to drive demand even higher

2

. Earlier this year, SK Hynix and partner Sandisk introduced HBF technology that stacks NAND flash circuits vertically, matching HBM performance while providing up to 16 times more capacity

4

.

Industry Braces for Worst Memory Supply Shortage in History

SK Hynix CEO Kwak Noh-jung has warned that 2027 will bring the most severe memory supply shortage the industry has ever experienced, with customer demand expected to outpace production capacity beyond 2030

3

5

. The company views current memory demand growth not as a temporary cycle but as a structural transformation where memory has become core AI infrastructure determining performance itself

3

5

.

Analysts project that both DRAM and NAND demand will maintain a compound annual growth rate of 19% from last year through 2030, according to Omdia data cited by SK Hynix

5

. The surge in memory prices driven by supply shortages and massive AI-driven demand has led to record financial performance for the biggest producers—SK Hynix, Samsung, and Micron—with their share prices rallying as investors bet the supply imbalance will persist

1

. SK Hynix reported record quarterly operating profit of 60.54 trillion won in the second quarter, a 557% year-on-year increase, with cumulative first-half revenue crossing 100 trillion won for the first time in company history

3

.

Strategic Positioning for AI Infrastructure Dominance

An SK Hynix official stated that "in the AI era, technological competitiveness alone is not enough and the ability to supply the required volume at the exact moment customers need it is the ultimate competitive advantage"

1

. The company reached this investment decision after thorough review of market demand and aims to establish itself as a key partner in AI infrastructure, contributing to supply chain stabilization

3

.

Source: TweakTown

Source: TweakTown

Memory manufacturers have faced criticism for alleged unwillingness to build new manufacturing capacity, with SK Hynix previously predicting the supply chain crisis would peak between 2027 and 2030

2

. The new chip fabs should increase production capacity substantially, though they're scheduled to come online just as the market may begin settling. However, the investment is clearly aimed at enterprise servers and the AI accelerator market rather than consumer applications, meaning near-term relief for consumer RAM prices remains unlikely

5

. This positions SK Hynix to maintain market share dominance in the AI memory segment as competition with Samsung intensifies across HBM, NAND, and emerging vertical memory technologies.

Today's Top Stories

© 2026 TheOutpost.AI All rights reserved