The End of the Jensen Huang Era and the Dawn of ‘AI 2.0’: The Collapse of NVIDIA’s Monopoly and New Opportunities for Samsung and SK Hynix

The resignation of founder Jensen Huang as CEO and his departure from the board after leading NVIDIA for 33 years is a monumental event that goes far beyond a simple executive reshuffle. It marks a structural watershed that will fundamentally reshape the capital circulation structure that has dominated the global AI ecosystem for years, the semiconductor supply chain (foundry and memory) spanning South Korea, Taiwan, and the United States, and the U.S. Congress’s overall semiconductor export control policies toward China.

To help general readers and investors clearly grasp the essence of this massive tectonic shift, we break down the fallout into four key issues—from Jensen Huang’s future trajectory and the tangible impact on Samsung Electronics and SK Hynix to the outlook for semiconductor fabs in Texas, USA.

1. Jensen Huang’s Future Trajectory: Why Acting as a ‘Shadow Architect’ Is Impossible

The very first question many people ask is, “Even though Jensen Huang has stepped down from his official titles, won’t he still pull the strings behind the scenes as a major shareholder to direct new AI chip development and management?” To put it bluntly, the likelihood of that happening is extremely low.

In a typical corporation, a founder might remain as an “honorary chairman” or “advisor” to exert behind-the-scenes influence. However, this situation is directly tied to national security and export control violation allegations (involving roughly $7.5 billion worth of illicit AI chip smuggling to China) that are under intense scrutiny by the U.S. Department of Justice, the Department of Commerce, intelligence agencies, and Congress.

• The Necessity of a Strict Management Firewall: For interim CEO Colette Kress (former Chief Financial Officer) and Chairman of the Board Harvey Jones, the top priority is preventing worst-case scenarios, such as NVIDIA being placed on U.S. government sanctions lists (like the Entity List) or facing crippling fines. To achieve this, they must transparently prove to regulators and shareholders that “NVIDIA is completely separated from founder Jensen Huang both operationally and technologically.”
• Realistic Future Scenarios: Compounded by the risk of shareholder class-action lawsuits alleging corporate value erosion due to personal dealings, Jensen Huang—who remains the fourth-largest shareholder with an approximate 3.61% stake—has no choice but to remain a Passive Shareholder focused on navigating legal risks and defending his asset value for the foreseeable future. His days of meeting global Big Tech leaders to orchestrate the industry are virtually over.

2. Contraction of NVIDIA’s ‘Circular Financing’ and Restructuring of Samsung & SK Hynix Supply Chains

Unlike Jensen Huang, an engineer by background who drove aggressive technological expansion and ecosystem dominance, interim leader Colette Kress is a rigorous financial and risk-management leader recognized across Wall Street and Silicon Valley. With this shift in leadership style, existing supply chain contracts and Memorandums of Understanding (MOUs) will also be realigned under the cold criteria of “technological necessity” and “financial soundness.”

Diminishing Role as the ‘AI Central Bank’ (The End of Circular Financing)

Until now, NVIDIA sustained explosive growth through so-called “Circular Financing”—investing astronomical sums into promising AI startups like OpenAI, Anthropic, and CoreWeave, which in turn used that invested capital to purchase NVIDIA’s high-priced GPUs. Simply put, NVIDIA acted as an “AI Central Bank” that both lent the money and sold the goods.

Under Kress’s leadership, which prioritizes financial soundness, brakes are highly likely to be applied to such high-risk capital circulation. Consequently, the growth rate of standalone GPU orders for NVIDIA may temporarily decelerate.

Windfall Gains for South Korean Semiconductor Giants (SK Hynix & Samsung Electronics)

A slowdown in NVIDIA’s solo run does not mean South Korean semiconductor companies will suffer a blow. Rather, as market power decentralizes, new doors of opportunity are opening.

1. SK Hynix (Maintaining Bargaining Power and Client Diversification): The biggest bottleneck in current AI computing speed lies not in the chip’s processing power, but in how fast data can be transported (memory bandwidth). Holding the de facto standard in the High Bandwidth Memory (HBM3E and HBM4) market, SK Hynix will maintain strong pricing power by expanding direct HBM supply not only to NVIDIA, but also to Big Tech giants like Google and Microsoft (MS) that are developing their own proprietary chips.
2. Samsung Electronics (Expanding Foundry Order Opportunities): Some comprehensive MOUs previously driven by personal trust or a top-down approach between Jensen Huang and Samsung’s top executives may be reviewed at the working level. Paradoxically, however, as Big Tech firms seek to step out from under NVIDIA’s shadow and accelerate the production of their own Application-Specific Integrated Circuits (ASICs), the strategic value of Samsung Foundry—equipped with both cutting-edge sub-nanometer processes and advanced packaging capabilities—is rising higher than ever.

3. Democratic Control of U.S. Congress and Utilization Outlook for Texas Semiconductor Fabs

With the Democratic Party taking control of Congress following the U.S. midterm elections, industrial policy and regulatory environments have also reached a major turning point. This directly impacts the utilization rates of U.S.-based semiconductor manufacturing facilities (Fabs), including Samsung Electronics’ Taylor plant in Texas.

• Short-Term Breather (Potential 1–2 Quarter Delay): Due to the leadership vacuum at NVIDIA, global AI data center capital expenditure (CapEx) execution may enter a temporary wait-and-see phase. As clients recalibrate chip order timelines, the initial full-scale utilization of new domestic U.S. foundry plants could be delayed by one to two quarters compared to original schedules.
• Explosive Shift in Mid-to-Long-Term Demand (‘Made in USA’ Security Premium): Over the long term, however, order books for U.S.-based fabs are projected to become even more robust. A Democratic-led Congress is highly likely to intensify antitrust investigations into Big Tech monopolies and circular financing, while simultaneously pushing legislation that mandates the use of domestically produced semiconductors (Trusted Foundry) for national security, defense, and public AI infrastructure.
• Big Tech Coalition Filling Texas Fabs: Furthermore, major cloud providers (hyperscalers) such as Amazon (Trainium), Google (TPU), and Meta (MTIA) are drastically increasing the mass production of their custom-designed AI chips to reduce reliance on NVIDIA. Ultimately, the long-term full-scale demand for advanced U.S. facilities in Texas and beyond will be filled seamlessly not by a single company like NVIDIA, but by custom chip orders from a coalition of Big Tech giants.

4. Comprehensive Outlook: The End of ‘AI Era 1.0’ Ruled by a Single Emperor

Jensen Huang’s departure signifies the official conclusion of ‘AI Era 1.0’, where a single charismatic founder and a lone corporation (NVIDIA) monopolized everything from capital financing to the hardware and software ecosystems.

The upcoming ‘AI Era 2.0’ in the global semiconductor market will be reshaped around three core pillars:

1. Diversification of Proprietary AI Chips (ASICs) by Big Tech: Moving away from a monolithic reliance on general-purpose GPUs, the era of custom chips tailored to each company’s specific services will begin in earnest.
2. Strict Export Controls and AI Regulation Led by the U.S. Congress: As national security and antitrust regulations tighten, the strategic value of domestic manufacturing (On-shoring) will be maximized.
3. Rising Valuation of Core Manufacturing and Memory Supply Chains: Regardless of who designs the chips, only a handful of companies worldwide can actually fabricate them (TSMC and Samsung Foundry) and supply ultra-high-speed memory (SK Hynix and Samsung Electronics). The more chip design power decentralizes, the stronger the strategic bargaining power of the core manufacturing and memory companies supporting them will become.

💡 Key Terms at a Glance for General Readers

• Foundry: A specialized semiconductor contract manufacturing plant that does not design chips itself, but produces them based on blueprints provided by client companies (e.g., Taiwan’s TSMC, South Korea’s Samsung Electronics).
• HBM (High Bandwidth Memory): An essential high-performance memory semiconductor for the AI era that vertically stacks multiple DRAM chips to dramatically boost data processing speeds.
• ASIC (Application-Specific Integrated Circuit): Unlike NVIDIA’s general-purpose AI chips, these are custom-designed AI semiconductors tailored by specific companies like Google or Amazon to optimize their own AI services.
• CapEx (Capital Expenditure): Large-scale investment funds spent by a company to build factories, construct data centers, or acquire physical assets to generate future profits.