- South Korea's dominance in AI-critical memory chips is making it an indispensable part of the global AI infrastructure supply chain
- Surging hyperscaler spending on AI data centers is boosting demand not only for GPUs but also for HBM, DRAM and NAND, where Korean firms lead globally
- While Korea is benefiting from the AI boom, growing dependence on memory chips also exposes the country to greater concentration and technology-cycle risks
As hyperscalers accelerate investments in AI infrastructure, South Korea has emerged as one of its biggest beneficiaries, not because it builds the world's largest AI models or operates the biggest data centers, but because it produces the memory chips that power them.
A new report by Morningstar-DBRS argues that the country's semiconductor industry has become increasingly central to the AI economy, with Korea's leadership in memory chips driving exports, corporate earnings and economic growth while simultaneously increasing its exposure to global technology cycles.
“While the semiconductor industry is broadly divided between logic chips and memory chips, Korea's strength lies in the latter, with domestic firms accounting for roughly 70% of global memory production and an estimated 80–90% of HBM [High Bandwidth Memory] output. The rapid adoption of AI since 2022 has reinforced this position, placing Korea at the center of the global AI value chain,” noted the Morningstar DBRS report.
"Korea sits at the center of the AI buildout because the memory AI systems depend on is largely a Korean product," Jeongku Choi, Research Analyst at Counterpoint Research told Fierce Network. "SK Hynix and Samsung together supply most of the world's High Bandwidth Memory, the stacked DRAM that feeds AI accelerators."
That shift marks a significant departure from the earlier semiconductor cycles, which were largely driven by smartphones and consumer electronics. Today's growth is increasingly being fueled by hyperscalers investing billions of dollars in AI infrastructure, making memory as strategically important as compute.
The AI boom extends well beyond GPUs
Memory has emerged as one of the defining technologies of the AI era, even if it has received far less attention than the GPUs powering today's generative AI models.
"The demand isn't limited to HBM or to GPUs. AI servers pair those accelerators with high-power CPUs, and each node also carries a lot of DRAM and NAND for the host system. So, the buildout is draining multiple ends of the memory market at once: premium HBM, mainstream server memory and NAND storage capabilities," said Choi.
This is reshaping South Korea's role within the semiconductor industry. "Korea's two makers [Samsung and SK Hynix] lead in all three [HBM, DRAM AND NAND]. Now that memory is a pivotal element in the AI ecosystem, Korea is no longer just a commodity memory supplier. It's now hard to build AI compute without going through it,” added Choi.
The Morningstar DBRS report notes that Korea's semiconductor industry has become increasingly aligned with global AI infrastructure spending, making hyperscaler capital expenditure a key driver of the country's economic performance.
And the scale of that investment continues to grow. According to Dell'Oro Group, AI expansion is likely to drive worldwide data center capex to $1.7 trillion by 2030. Prominent tech companies, including Microsoft, Amazon, Google and Meta have collectively committed hundreds of billions of dollars toward AI infrastructure over the next several years.
Memory has become strategic infrastructure
HBM is crucial because it enables GPUs to process massive AI datasets by providing significantly higher bandwidth than conventional memory technologies. Without HBM, many of today's large language models would struggle to achieve current performance levels. This has led to the growth of companies such as SK Hynix and Samsung from traditional semiconductor suppliers to strategic partners within the AI ecosystem.
The Morningstar DBRS report notes that Korea's dominance in memory semiconductors has positioned it among the principal beneficiaries of expanding AI investment, supporting exports, corporate profitability, government revenues and broader economic growth.
The country's export profile increasingly reflects that shift. “Semiconductor exports recorded an average growth of 32% during 2024-2025 and was 60% in 1H 2026. Semiconductors now account for roughly 30–35% of Korea's total exports. Although Korea remains a diversified manufacturing economy with sizeable automobile, shipbuilding, and petrochemical industries, recent growth has become increasingly semiconductor led,” says the Morningstar DBRS report.
Success brings new risks
However, this growth in South Korea’s profile in the semiconductor industry is not without threats. As AI investment becomes the primary driver of memory demand, Korea's economy is becoming increasingly exposed to fluctuations in hyperscaler capital spending and broader technology investment cycles.
“The sector's capital-intensive nature may amplify cyclical volatility, while Korea's reliance on external demand, particularly from US hyperscalers and global technology firms increases its sensitivity to changes in global technology investment cycles. Concentration risk is compounded by Korea's specialization in memory semiconductors, as it leaves the country exposed to future technological shifts, changing AI architectures, or industrial policies in other economies,” says the report.
In addition, the country’s financial markets are also becoming closely tied to the fortunes of the semiconductor industry. With Samsung and SK Hynix accounting for a significant share of the country’s stock market indices, investor sentiment is increasingly influenced by developments in the global chip sector. This can have significant ramifications when the demand fluctuates or starts to drop.
As AI infrastructure spending accelerates, demand for HBM, DRAM and NAND is expected to rise alongside investments in networking and power infrastructure. While South Korea has emerged as a key player in the AI ecosystem, it needs to protect itself from the vulnerabilities of overexposure to fluctuations in capital spending and technology investment cycles.
Read Fierce Network's related coverage
Memory shortage puts the squeeze on telco modernization efforts
Samsung charts its own path to AI-RAN
Korean operators ramp up AI investments to make country an AI superpower