AI Chips, Export Controls, and a New Phase in US–China Tech Tensions
US–China technology and trade tensions are entering a more complex phase as an unprecedented boom in artificial intelligence (AI) infrastructure runs headlong into tightening export controls on semiconductors and critical raw materials. Over the past few days, several developments have underscored how deeply this struggle now intersects with corporate earnings, capital expenditure plans, and supply-chain resilience for US businesses.
On the one hand, AI demand is driving a global shortage in key components such as high-end GPUs and memory chips, pushing prices higher and triggering a new investment cycle across data centers, cloud computing, and networking. On the other, Washington is expanding export controls on advanced semiconductor technology to China, while Beijing continues to wield its dominance in critical raw materials as a strategic tool.
Recent commentary from Nvidia, emerging analysis of the memory market, and new research on China’s control of critical raw materials highlight how this tug-of-war is reshaping the operating environment for US firms from Big Tech to industrials and automakers.
Nvidia’s $200 Billion CPU Ambition Still Includes China
Nvidia remains at the epicenter of the AI infrastructure boom. During its latest earnings call last week, CEO Jensen Huang said the company’s new "Vera" central processing units (CPUs) open access to what he estimates is a $200 billion addressable CPU market tied to AI and data center workloads. Over the weekend, Huang clarified that this forecast explicitly includes China, even as US export controls restrict the sale of Nvidia’s most advanced AI accelerators into the market.
Speaking at an industry event, Huang reiterated that Nvidia sees China as a long-term AI and data center opportunity, but emphasized the company is working to navigate export regulations for its advanced chips. The remarks signal two things for investors and corporate decision makers:
First, Nvidia is not writing off China; it is attempting to design and position products that fall within the evolving regulatory perimeter while still capturing demand from Chinese cloud, internet, and enterprise customers.
Second, the company is simultaneously deepening its supply-chain ties with non-mainland hubs, including continued investments in Taiwan’s chip ecosystem, to secure capacity for the broader global market.
For US markets, Nvidia’s stance underscores that export controls are a constraint, not an absolute barrier. However, the policy overhang introduces a persistent valuation and earnings risk premium — not only for Nvidia, but for the broader US semiconductor complex and the companies that depend on it.
A Global Memory Chip Supply Crunch, With China Waiting in the Wings
The AI wave is no longer just a GPU story. According to a report in The Wire China on May 24, the world is facing a global shortage of memory chips, driven primarily by a surge in demand from AI data centers. Industry executives described the shortage as “unprecedented in scale,” as generative AI workloads require dramatically higher memory capacity and bandwidth per server than traditional cloud applications.
At the same time, the United States has expanded export controls on semiconductor manufacturing equipment sales to Chinese companies over the past several years. These restrictions are designed to slow China’s progress in high-end chip fabrication — including advanced DRAM and NAND — by limiting access to critical tools from US, Japanese, and Dutch suppliers.
Yet Chinese memory players, backed by substantial state support, are pushing aggressively to fill the gap. The Wire China analysis highlights how China’s expanding role in memory production is increasingly being viewed — by some in global industry — as a potential partial answer to the current supply crunch, especially for commodity or slightly trailing-edge memory where export restrictions are less binding.
This dynamic has several implications for US businesses:
Higher input costs for AI builders: US cloud providers and hyperscalers, already paying premium prices for GPUs, are now facing rising costs and potential allocation constraints for DRAM and NAND. This can compress margins or force higher pricing for AI services.
Strategic sourcing decisions: US firms may be tempted to lean more heavily on Chinese memory suppliers to secure capacity, but doing so increases exposure to future policy shifts and sanctions risk.
Capex and inventory swings: The sharp tightness in memory could trigger a new wave of capacity expansion by Korean, US, and potentially Chinese producers. Over time, that raises the risk of another boom-bust cycle in pricing, with downstream consequences for hardware OEMs and data center operators.
China’s Critical Raw Material Leverage Remains a Structural Risk
While the US is tightening controls on chip technology, China retains considerable leverage over the upstream inputs that make high-tech manufacturing possible. A new policy brief from the EU Institute for Security Studies, released in May 2026, underscores the scale of Beijing’s dominance across critical raw material supply chains.
According to the report, China controls over 74% of global production in at least one value chain segment for each of the key strategic materials studied, and for heavy rare earths and gallium, its share reaches as high as 98% or more as of 2025. The study notes that throughout 2025, China used export controls on certain critical materials to gather strategic information, pressure trading partners, and erode their industrial capabilities, including in defense-related sectors.
For US businesses, this concentration represents a structural vulnerability, particularly in sectors such as:
Semiconductors and electronics: Gallium and germanium are essential for advanced chips, power electronics, and certain photonics applications.
EVs and clean energy: Rare earths are crucial for high-performance permanent magnets in electric motors and wind turbines.
Defense and aerospace: Specialty alloys and high-performance materials rely on rare earths and other critical elements.
Any escalation in US–China tech tensions that prompts Beijing to tighten export quotas or impose new licensing requirements could quickly reverberate through US manufacturing supply chains, increasing costs, delaying production, and forcing expensive diversification efforts.
How US Corporates Are Being Reshaped
1. Earnings and Margin Dynamics in Tech and Semiconductors
US semiconductor firms are benefiting from exceptionally strong AI-related demand, but are simultaneously contending with policy, capacity, and cost constraints:
Revenue concentration: AI data center spending is increasingly concentrated among a handful of hyperscalers and large enterprises, elevating customer concentration risk even as revenues surge.
China revenue exposure: For Nvidia and peers, the ability to sell downgraded or compliant products into China remains critical to sustaining growth. Any tightening of US rules that further restricts these sales could cut into forward earnings forecasts.
Cost inflation from memory and materials: As memory prices rise and critical materials remain strategically controlled, chipmakers face upward pressure on bill-of-materials costs, which may not be fully pass-through in the short term.
For broader US technology companies — cloud providers, enterprise software vendors, and device makers — the implication is a mixed but generally positive earnings picture. AI infrastructure build-out supports revenue and services growth, but higher capital intensity and component costs can weigh on near-term margins.
2. Capital Expenditure and Supply-Chain Rewiring
US businesses across sectors are accelerating capital spending in response to these tensions, but the direction of that spending is shifting:
Onshoring and friend-shoring: Incentivized by US industrial policies like the CHIPS and Science Act and by geopolitical risk, companies are looking to expand or secure semiconductor capacity in the US, Taiwan, South Korea, and allied markets rather than relying on Chinese fabs.
Alternative sourcing of critical materials: Mining and processing projects in North America, Australia, and Europe are drawing greater strategic interest and investment as customers seek to reduce exposure to Chinese-controlled supply chains.
Inventory buffers: To mitigate shock risk, firms in sectors such as autos, industrial equipment, and defense are rethinking just-in-time models and building larger inventories of critical chips and components.
These shifts are capital-intensive. For many corporates, operating cash flow is being increasingly redirected toward resilience-oriented investment — new suppliers, redundant capacity, and long-term offtake agreements — which can support growth over the long term but may limit buybacks or dividends at the margin in the near term.
3. Strategic Risk Management and Policy Monitoring
US companies now treat US–China tech policy as a core strategic variable rather than an externality. Boards and executive teams are:
Building scenario analyses around potential further restrictions on chip exports, cloud services, or data flows.
Assessing the risk that China responds to US actions with targeted export controls on critical raw materials or components.
Reevaluating joint ventures, R&D cooperation, and capital deployment in China, balancing growth opportunities against regulatory and reputational risks.
This is particularly relevant for US companies with large China exposures in sectors where technology, data, or dual-use capabilities intersect with national security concerns.
Broader US Economic Implications
AI-Driven Growth vs. Policy Friction
On a macro level, the AI boom is a significant positive for US growth potential. Spending on AI-related hardware, software, and services is filtering down through the economy via increased demand for construction, power infrastructure, fiber networks, and professional services.
However, the friction from US–China tech tensions tempers this upside in several ways:
Higher capital costs: Redundancy and diversification raise the cost of doing business relative to a world with unfettered global trade.
Persistent inflationary pressure in key inputs: Semiconductors, critical materials, and specialized equipment may remain structurally more expensive due to fragmented supply chains and strategic stockpiling.
Uncertain productivity payoffs: If policy frictions slow diffusion of AI technologies into manufacturing and services, the expected productivity gains could arrive more gradually.
Manufacturing and Industrial Rebalancing
US manufacturing stands to gain selectively from the recalibration of supply chains. New domestic and allied capacity for chips, EV components, and advanced materials can support jobs and regional development. Yet these gains are uneven and may be offset by higher input costs and retaliatory measures in other sectors.
For industrial conglomerates, aerospace and defense firms, and automakers, the key risk is not demand destruction but episodic disruption — sudden shortages of a key chip or material that delay product launches or production schedules. The experience of 2020–2022 semiconductor shortages remains fresh, shaping management attitudes toward risk and resilience.
What Investors and Corporates Should Watch Next
Given the latest developments, there are several key signposts that will determine how strongly US–China tech tensions feed into US corporate earnings and the broader economy:
Further US export control actions: Any expansion of restrictions on AI chips, semiconductor tools, or cloud services to China would have immediate implications for revenue guidance at leading US tech and semiconductor firms.
Chinese policy responses: New or expanded export controls on rare earths, gallium, or other critical materials would directly impact US electronics, EV, and defense supply chains.
Memory market balance: Whether the current memory chip shortage eases via new capacity or worsens due to additional demand will shape hardware and data center cost curves into 2027.
Corporate capex signals: The pace and geographic allocation of AI and semiconductor-related investment by US firms will indicate how they are internalizing policy risk and supply constraints.
Conclusion: A Bullish Structural Trend with Elevated Risk Premiums
The intersection of AI infrastructure demand and US–China tech tensions is now one of the defining forces in global business and markets. Nvidia’s inclusion of China in its $200 billion CPU market ambition highlights the enduring scale of Chinese demand, even under constraints, while the global memory shortage and Beijing’s dominance in critical raw materials underscore how tightly intertwined the two economies remain.
For US businesses, the net effect is a powerful but more complex growth story. AI and advanced computing are likely to support strong top-line expansion across semiconductors, cloud, and related industries, and to catalyze investment in domestic and allied manufacturing. At the same time, export controls, strategic materials leverage, and geopolitical uncertainty are pushing up the cost of resilience and embedding a higher risk premium into earnings and valuations.
In this environment, companies that can secure diversified supply chains, navigate regulatory constraints, and convert AI demand into scalable, high-margin products and services are poised to outperform. For investors and corporate leaders alike, closely tracking the evolving policy landscape on both sides of the Pacific is no longer optional — it is central to understanding the next phase of US growth, profitability, and industrial strategy.
