Nvidia and Intel navigate a volatile 2025 for US chips

The U.S. semiconductor industry experienced a tumultuous 2025, characterized by rapid shifts in export controls (government restrictions on selling technology abroad), major leadership shakeups, and complex trade negotiations with China. The volatility began early in the year. On February 28, Intel delayed construction on its $28 billion semiconductor project in Ohio. Seeking a turnaround, Intel named Lip-Bu Tan as its new CEO on March 12. According to Intel CEO Lip-Bu Tan, the manufacturer would operate with a focus on engineering under his leadership. Meanwhile, corporate consolidation and regulatory pressures mounted. On April 3, Intel and TSMC allegedly reached a tentative agreement to launch a joint chipmaking venture. Shortly after, on April 15, the Trump administration restricted chip exports, leading Nvidia to expect $5.5 billion in charges related to the new requirements.

By mid-year, policy shifts and corporate restructuring intensified. On May 7, the Trump administration planned on taking a different path regarding the Framework for Artificial Intelligence Diffusion, ultimately rescinding the policy. These shifting export controls heavily impacted balance sheets; on May 28, Nvidia reported that licensing requirements cost the company $4.5 billion in Q1 charges, with an expected $8 billion revenue hit in Q2. In response to financial pressures, Intel began offloading non-core assets, including its Network and Edge units, which had generated $5.4 billion in 2024 revenue. Intel also began laying off 15% to 20% of workers in its foundry business unit, part of a broader pullback to end the year with around 75,000 employees.

Late summer brought geopolitical friction and leadership crises. On July 17, a groundbreaking Middle East AI deal was reportedly on hold due to U.S. national security concerns. On August 7, President Donald Trump demanded that Intel’s CEO resign immediately, citing conflicts of interest. Amid the turmoil, SoftBank took a $2 billion stake in Intel on August 18. To maintain market access, Nvidia and AMD struck a deal on August 12, agreeing to pay the U.S. government 15% of revenue from China sales. However, regulatory hurdles remained; on September 15, China’s State Administration for Market Regulation ruled that Nvidia violated the country’s antitrust regulations (competition law).

The year concluded with major product and financial milestones. On October 9, Intel announced its Panther Lake processor, to be built exclusively at its Arizona fab (semiconductor fabrication plant). On November 19, Nvidia reported Q3 revenue of $57 billion, a 66% increase over 2024. On December 8, the U.S. Department of Commerce decided that Nvidia and AMD can send AI chips to China. Finally, on December 24, Nvidia struck a non-exclusive licensing deal with Groq, purchasing $20 billion of its assets.

Why it matters

The U.S. semiconductor industry experienced significant volatility in 2025, marked by shifting export controls, leadership changes at major firms like Intel, and evolving trade relations with China. These dynamics underscore how geopolitical competition and regulatory policy directly dictate the commercial boundaries of the hardware supply chain.