The semiconductor industry is more fragmented than one may know.
U.S.-based semiconductor companies dominate headlines with a 50.4% global market share, but the factories that make the chips are mostly in Asia. Wall Street rewards Nvidia Corp. (NASDAQ: NVDA), Broadcom Inc. (NASDAQ: AVGO) and Advanced Micro Devices Inc. (NASDAQ: AMD) for their soaring revenue and design IP, yet Taiwan’s TSMC (NYSE: TSM), South Korea’s SK Hynix (KS: 000660), and Malaysia’s OSAT firms quietly power the machines that turn those designs into silicon.
That geographic separation doesn’t just cause confusion over who gets the credit. It creates financial and policy distortions around how revenue is booked, where capital is deployed, and how tariffs are levied. These issues are followed closely by the Gilder’s Technology Report, including an in-depth analysis in its latest update.
Consider modern smartphones. A chip design may originate at Apple (NASDAQ: AAPL) in Cupertino, California. But the product is fabricated in Taiwan at TSMC. From there, the finished wafers head to Malaysia for back-end packaging. Then they’re sent to China, where final device assembly—including the printed circuit board, camera module, and screen—is done by Foxconn or another engineering, manufacturing and services. When that smartphone is finally shipped back to the United States, tariffs are applied—not at the chip level or PCB level, but on the entire assembled device.
This mismatch has three cascading implications:
Chart 1

Semiconductor Market Share by Headquarters vs. Sales Geography
U.S.-headquartered semiconductor firms account for more than half of global revenue, reflecting their dominance in chip design, systems architecture, and branded platforms (see Table 1). Korea follows with 21.1%, driven largely by memory giants Samsung and SK hynix. But what’s critical here is that this chart tracks the location of company headquarters, not the location of fabs, tooling, or capital spending. It is a map of where value accrues on paper, not where silicon gets produced.

The World Semiconductor Trade Statistics (WSTS) dataset provides a different lens—semiconductor revenue by geographic region of sale (see Table 2). Only 29.8% of global semiconductor purchases happened in the Americas in 2024. Asia-Pacific accounted for over half. This highlights a critical disconnect between where chips are sold or consumed versus where design IP is headquartered. It also reinforces how globalized the industry remains: U.S. firms may control the architecture, but most end-markets—and therefore geopolitical exposure—exist outside the United States.

Manufacturing Power Revealed Through Equipment Imports
As Table 3 shows, semiconductor equipment imports in 2024 were overwhelmingly concentrated in Asia, with China alone accounting for nearly $50 billion in tool purchases. Korea and Taiwan followed closely behind, while North America trailed with just $13.7 billion. These figures tell a deeper story: although U.S. firms dominate the chip design narrative, the physical scaling of semiconductor manufacturing remains firmly rooted in East Asia. This table exposes the true geography of capacity growth and capital deployment—a reality that supply chain policies must confront.

Valuation Favors Design, but Manufacturing Holds the Leverage
The market capitalization leaderboard as of May 30, 2025, continues to highlight the divergence between IP ownership and physical production (see Table 4). U.S. firms dominate the top rankings—Nvidia at $2.476 trillion, Broadcom at $804 billion—while the key fabrication and tooling players like TSMC, Samsung, and ASML remain undervalued by comparison. This valuation gap reflects investor focus on scalability and software leverage, but masks how irreplaceable the manufacturing layer has become in a geopolitically fractured world.

Log In
Forgot Password
Search