📊 Full opportunity report: The Hidden Chokepoint In AI Growth: Seoul Addresses Memory on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
SK hynix’s chairman warns of an imminent AI memory shortage due to lack of new capacity, with demand expected to outpace supply by 50-60% in 2027. This shortage has geopolitical implications and affects AI infrastructure costs.
SK hynix’s chairman, Chey Tae-won, has publicly warned that the global demand for AI memory will significantly outpace supply in 2027, with an increase of 60 to 100 percent over current levels, and no meaningful new capacity coming online before then. This development raises concerns about a looming memory shortage that could impact AI deployment, geopolitical stability, and semiconductor markets.
During a press briefing at the Korea Chamber of Commerce and Industry’s Jeju Forum, Chey Tae-won highlighted that AI now accounts for more than half of total semiconductor consumption, with demand growth expected to be at least 50-60 percent in 2027. Despite this, he stated that no significant new capacity is scheduled to be operational in 2026 or early 2027, creating a supply shortfall.
The shortage is most acute in high-bandwidth memory (HBM), which is critical for AI accelerators. SK hynix currently holds 58 percent of the global HBM revenue, with Micron and Samsung sharing the rest. The company has announced capacity expansions, including a new plant expected to start in February 2027, but these will not meet the immediate demand spike.
Chey also warned that high memory prices, driven by supply constraints, are leading to ‘chipflation,’ which increases costs for device makers and could provoke geopolitical responses. He emphasized that the current pricing environment is abnormal and unsustainable, and that the industry must address this bottleneck to avoid further market instability.
Models get the headlines.
Memory is the chokepoint.
SK Group’s chairman at the Jeju Forum, per The Korea Herald: customers want 60–100% more AI memory in 2027, governments now treat memory access as economic security — and no company has meaningful new capacity arriving next year.
The gap, in his own numbers
customer requests to SK hynix vs this year. AI already consumes over half of all semiconductors; total demand growth floored at 50–60%.
“No company has meaningful new capacity coming online next year.” The gap year is already locked in — fabs don’t move faster than physics.
Result, per Chey: near-chaotic lobbying — no longer just from companies. Foreign governments are intervening for domestic industries; next, governments pressure governments.
Tighter than the chokepoints you worry about
SK hynix’s race against its own warning
Company figures and projections as announced — none of it lands in 2026.
Half true: unified-memory Apple Silicon doesn’t queue for HBM — a fleet you own is insulated from allocation politics, and owned hardware converts supply-chain risk into sunk cost.
The other half: LPDDR and HBM share DRAM wafer economics — chipflation reaches workstation memory too, and training compute stays fully hostage. Local inference changes who feels the shortage, not whether it exists.
Week tie-in: if memory demand grows into capacity that doesn’t exist, doing the job in 3B parameters on memory you already own isn’t aesthetics — it’s engineering under constraint.
MEMORY WAR: HBM's Dominance Beyond NVIDIA — The 12-Year Monopoly Formula (The Memory Hegemony Series Book 1)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Implications of Memory Shortage on AI and Geopolitics
The warning from SK hynix’s chairman underscores a critical bottleneck in AI infrastructure development, with potential ripple effects across global markets and geopolitical relations. As demand for AI accelerators and high-bandwidth memory grows rapidly, the lack of capacity could lead to increased costs, supply chain disruptions, and heightened geopolitical tensions, especially as governments treat memory access as a matter of economic security.
This situation highlights the risks of industry concentration, with SK hynix holding a majority share of the HBM market, and the potential for supply-side vulnerabilities to influence global AI competitiveness and security policies.
Memory Capacity and Geopolitical Market Dynamics
The current semiconductor landscape is characterized by a tight oligopoly, with SK hynix controlling 58 percent of the global HBM revenue in Q1 2026, and Samsung and Micron each holding about 21 percent. Despite ongoing capacity investments, the industry faces a significant gap between demand and supply, especially as AI adoption accelerates.
Previous developments include SK hynix’s announcement of a $12.9 billion investment in a new packaging plant in Cheongju, aiming for a 33 percent CAGR in HBM demand through 2030. However, these capacity increases will not materialize until 2027, leaving a critical shortfall in the immediate future.
Meanwhile, geopolitical considerations are intensifying, with governments increasingly viewing memory access as a strategic asset, adding pressure to the supply chain and potentially leading to export controls or interventionist policies.
“No company has meaningful new capacity coming online next year.”
— Chey Tae-won, SK Group Chairman
Unresolved Aspects of the Memory Shortage Crisis
It is not yet clear how quickly capacity expansions will be completed or whether new capacity can be accelerated given current investment timelines. The full geopolitical response and potential market interventions remain uncertain, as do the precise impacts on AI deployment and pricing in the near term.Next Steps for Industry and Policy Response
Industry players will likely accelerate existing capacity projects and explore alternative memory architectures. Policymakers may begin to address strategic stockpiling or export controls, especially in regions where memory supply is concentrated. Monitoring the progress of SK hynix’s capacity expansions and global demand trends will be critical in assessing how the supply shortfall evolves.
Further announcements from SK hynix and other major memory suppliers are expected in the coming months, alongside potential policy measures from governments concerned about economic security and technological sovereignty.
Key Questions
How will the memory shortage affect AI development?
The shortage is likely to increase costs and slow deployment of large-scale AI models, especially those requiring high-bandwidth memory like HBM. This could limit progress in AI research and commercial applications.
Why is there no immediate capacity expansion planned?
Major capacity investments are scheduled for 2027, with existing projects already committed. Building new fabs takes years, and current investments do not meet the rapid growth in demand.
Could geopolitical tensions worsen due to this shortage?
Yes, as governments view memory access as a strategic asset, the supply constraints could lead to export restrictions, tariffs, or other measures to secure domestic industries.
What can companies do to mitigate this risk?
Companies can stockpile existing hardware, optimize memory usage, diversify supply sources, and invest in alternative architectures to reduce dependence on HBM capacity.
Will prices for consumer memory also increase?
High memory prices driven by supply constraints are already affecting consumer and workstation memory markets, and further increases are likely if shortages persist.
Source: ThorstenMeyerAI.com
