Why It Matters
Intel’s trajectory has broader implications than its own balance sheet. As the Trump administration moves to rebuild domestic semiconductor manufacturing, Intel sits at the center of a national strategy to reduce dependence on foreign chip production — a goal with economic and security stakes for the entire country.
What Happened
Once the unquestioned king of American chip manufacturing, Intel spent years watching rivals chip away at its dominance. Nvidia captured the AI training market with graphics processors, while AMD and Qualcomm made inroads in other areas. Meanwhile, Taiwan Semiconductor Manufacturing Company, known as TSMC, now produces roughly 90 percent of the world’s most advanced chips — a concentration that has rattled policymakers in Washington.
Into that difficult picture stepped Lip-Bu Tan, the 66-year-old Malaysia-born executive who took over as Intel’s chief executive officer in March 2025. Tan arrived with a reputation built during more than a decade leading Cadence Design Systems, and he wasted little time making his mark. He reduced Intel’s workforce by approximately 34 percent, paused plans to expand manufacturing facilities in Germany and Poland, and sold off a controlling stake in a subsidiary to shore up liquidity.
Tan also restructured the company internally, elevating engineers while trimming layers of middle management. He brought in senior executives recruited from Qualcomm and Arm to lead Intel’s data center and artificial intelligence divisions, and secured investment commitments from both Nvidia and Softbank. In a separate move with significant political weight, the Trump administration took roughly a 10 percent stake in Intel through an $8.9 billion investment, a move aimed at strengthening domestic advanced-chip production.
Last week, Tan delivered a keynote at Computex in Taipei, the annual trade show that has grown into one of the technology industry’s most-watched events amid the global AI boom. His message: the central processing unit — the CPU — is no longer a commodity afterthought. It is becoming essential infrastructure for the next phase of artificial intelligence.
The CPU Opportunity
The emerging case for CPUs centers on a distinction between two phases of AI work. Graphics processing units, or GPUs, dominate the training stage — the computationally intensive process of building AI models. But inference, the phase in which a trained model actually answers questions or performs tasks, increasingly relies on CPUs. That distinction matters as so-called agentic AI systems — programs that complete tasks on behalf of users — proliferate across platforms from OpenAI, Anthropic, and Google.
One example gaining traction is OpenClaw, an AI agent that operates apps and controls smart home devices through messaging applications. Systems like it depend heavily on inference computing, where Intel’s CPU lineup could find renewed relevance.
Notably, Nvidia — whose GPU dominance powered its rise to become one of the world’s most valuable companies — announced at Computex that it is pushing into the CPU market itself, with its Vera data center CPU now in full production. The move signals that the boundary between GPU and CPU workloads is shifting, and that Intel is not alone in recognizing the opportunity.
Dan Nystedt, vice president at investment firm TriOrient, framed the stakes plainly: “The CPU revival could save the company. The majority of Intel’s business is CPUs.”
By the Numbers
- 90% — Share of the world’s most advanced chips produced by TSMC
- 34% — Portion of Intel’s workforce reduced under Tan’s leadership
- $8.9 billion — Trump administration investment in Intel stock
- 10% — Approximate ownership stake acquired through that investment
- 300% — Surge in Intel’s share price since the August investment
Challenges Ahead
Despite the stock rally and the strategic pivot, Intel faces unresolved problems. Its foundry business — which manufactures chips for outside clients — has struggled to attract new customers and maintain consistent production quality. Tan has been candid about the scale of the work remaining. In remarks from April 2025, he acknowledged: “There are areas we need to improve, and there are no quick fixes.”
The broader jobs and economic picture provides some tailwind — a stronger economy tends to support technology investment — but Intel’s recovery remains a long-term project with no guaranteed outcome.
What’s Next
Tan’s near-term priorities include stabilizing the foundry operation, integrating the newly hired executives into Intel’s AI and data center divisions, and demonstrating that the CPU strategy can translate into tangible revenue gains. With the Trump administration’s stake now on the books and domestic chip policy drawing bipartisan attention in Washington, Intel’s performance carries weight beyond its own shareholders — making the next several quarters closely watched by both Wall Street and policymakers.
About the Author
