Technology · Dev
SK Group Chair Predicts Severe Memory Shortage by 2027
Tae-Won Chey acknowledges rapid price increases are driving chip inflation as demand surges across the semiconductor industry

KEY TAKEAWAYS
- ·SK Group Chairman Tae-Won Chey predicts the most severe memory shortage will occur in 2027 as explosive demand growth outpaces production capacity expansion.
- ·Memory chip prices have climbed sharply, with DDR5 modules rising approximately 40 percent since early 2025 and high-bandwidth memory used in AI accelerators more than doubling year-on-year.
- ·South Korea supplies roughly 70 percent of global DRAM, making regional supply concentration a key risk factor as the projected shortage approaches in 2027.
A Tight Market Ahead
The global memory chip market faces its most severe supply crunch in 2027, according to SK Group Chairman Tae-Won Chey. Speaking in a recent interview, Chey pointed to explosive demand growth that is outpacing the industry's ability to expand production capacity quickly enough to meet customer needs.
The warning comes as memory chip prices have climbed steeply over the past year, driven by surging consumption from artificial intelligence data centers, smartphone manufacturers, and cloud infrastructure providers. SK Hynix, the group's flagship semiconductor subsidiary, ranks among the world's top three memory producers alongside Samsung and Micron.
Chey acknowledged the strain that rapid price increases are placing on customers across the technology ecosystem. He offered an apology to the market for the pace of price hikes, recognizing that the swift escalation is contributing to broader chip inflation throughout the industry supply chain.
Capacity Constraints Meet Structural Demand
The projected shortage reflects a structural mismatch between long production lead times and accelerating consumption patterns. Memory fabrication plants require two to three years from groundbreaking to volume production, while demand for high-bandwidth memory and advanced DRAM has doubled in some segments over the past eighteen months.
Asian semiconductor manufacturers have announced capacity expansion plans totaling over USD 200 billion across the next five years, but the bulk of new production lines will not reach full output until 2028 or later. This timing gap creates the supply bottleneck Chey described.
The memory market has historically moved in cycles, with periods of oversupply driving price crashes followed by capacity discipline that tightens availability. The current cycle differs in its demand drivers. Previous shortages stemmed primarily from consumer electronics and PC upgrades, while today's consumption is anchored in enterprise infrastructure that requires continuous, predictable supply.
Price Pressure and Strategic Responses
Spot prices for DDR5 modules have risen approximately 40 percent since the beginning of 2025, according to industry tracking data. High-bandwidth memory used in AI accelerators has seen even steeper increases, with some contract prices more than doubling year-on-year.
Chey's public acknowledgment of price-driven inflation is unusual for a major supplier. The comments suggest SK Group recognizes that unchecked price increases risk damaging long-term customer relationships and may prompt buyers to seek alternative architectures or delay equipment upgrades.
The chairman's remarks also carry implications for capacity investment decisions. Memory producers have traditionally expanded production in response to sustained price strength, but the capital intensity of leading-edge fabrication plants and geopolitical pressure to diversify manufacturing locations complicate expansion planning.
Regional Implications
The looming shortage has particular significance for Asia's technology hubs. South Korea, home to Samsung and SK Hynix, supplies roughly 70 percent of global DRAM and more than half of NAND flash memory. Taiwan and Japan host additional memory production, but regional concentration means any supply disruption or demand surge ripples quickly through electronics manufacturing clusters in Shenzhen, Penang, and Ho Chi Minh City.
China's domestic memory producers have ramped production over the past three years, but output remains a fraction of established players and focuses on older technology nodes. Beijing's push for semiconductor self-sufficiency has yet to meaningfully alter the global memory supply balance.
Policymakers across the region are watching memory supply dynamics closely. Shortages drive up costs for downstream manufacturers, compress margins for electronics assemblers, and can delay product launches. Governments in Singapore, Malaysia, and Vietnam have offered incentives to attract memory and semiconductor investment, though few new fabrication plants have broken ground outside the traditional Korea-Taiwan-Japan triangle.
What Comes Next
The industry will test whether long-term supply agreements can smooth the volatility Chey described. Several major cloud providers and hardware manufacturers have shifted toward multi-year offtake contracts that guarantee volume in exchange for capacity commitments, a model borrowed from the foundry sector.
If the 2027 shortage materializes as projected, memory will join a growing list of semiconductor categories where demand has structurally outpaced supply. The question facing the industry is whether pricing discipline and capacity expansion can converge before shortages force customers to redesign systems around constrained components.
RELATED STORIES
Spot something wrong? Email editor@briefasia.com. We log every correction publicly.


