Intel shares have tripled since Lip-Bu Tan took charge. The stock set records in April. Yet the veteran chip executive has spent more time courting presidents, billionaires and device makers than overhauling the company’s aging factories.
Tan became chief executive in March 2025. For the first year he focused outward. He turned a tense White House exchange with Donald Trump into a major federal investment. The U.S. government now ranks as Intel’s third-largest shareholder. He forged a surprise factory partnership with Elon Musk. And preliminary talks with Apple signal potential manufacturing wins on American soil.
But inside the company the picture looks different. More than a dozen current and former employees described a CEO who has yet to spell out a detailed fix for Intel’s core weaknesses. Products still lag. Yields trail industry leaders. And a decade of manufacturing missteps won’t vanish because of strong handshakes.
The gap matters. Intel once owned nearly the entire data-center processor market. It held 99 percent share. Today it fights for every order on price, quality and speed. Relationships buy time. Execution decides survival.
Tan’s Network Delivers Capital and Credibility
The federal stake arrived in August 2025. An $8.9 billion investment gave the government a 9.9 percent holding, now valued near $36 billion after the share surge. The deal blended CHIPS Act support with a secure-chip program. Tan had called on Michael Dell and others to back him during earlier talks with Trump. (The Next Web, May 8, 2026)
Then came the Musk connection. Personal conversations led to Terafab, a proposed Texas chip complex. Initial costs could hit $55 billion. The total might reach $119 billion. Tesla intends to tap Intel’s upcoming 14A process. xAI’s AI5 chip sits among early targets. The arrangement caught many Intel leaders off guard.
Apple interest surfaced even more recently. The iPhone maker held early discussions about shifting some main device processors to U.S. production at Intel and Samsung. Such a move would mark the biggest validation yet for Tan’s foundry ambitions. Apple relies almost entirely on TSMC today. Any shift would reshape supply chains. (Bloomberg, May 5, 2026)
These wins fueled the stock. April brought a 114 percent jump, Intel’s best month in 55 years on Nasdaq. One session delivered a 24 percent gain, the strongest single-day move since 1987. First-quarter 2026 revenue hit $13.6 billion, beating forecasts by more than $1 billion. Data-center and AI sales rose 22 percent.
Optimism spread. Nvidia’s Jensen Huang spoke publicly about CPUs retaining a key role in AI systems alongside GPUs. That comment alone eased fears that Intel had missed the boom.
But Tan himself set the bar clearly. “Intel has the technology, talent and scale to lead again, but leadership is earned through execution,” he said in his first interview as CEO. (The Next Web, May 8, 2026)
He aims to finish recruiting a trusted internal leadership team by the end of June. The clock ticks.
Manufacturing shortfalls remain stark. Intel’s cost per chip runs roughly three times higher than TSMC’s. More than 40 percent of that gap traces to yields. Intel hovers near 65 percent. TSMC exceeds 80 percent. Only 8 percent of the difference comes from higher U.S. labor expenses. (The Next Web, May 8, 2026)
The 18A process sits at the center of hopes. It brings gate-all-around transistors and backside power delivery. Intel began high-volume manufacturing in late 2025. Panther Lake mobile chips and 288-core Clearwater Forest server processors rely on it. Yet competitive yields may not arrive until 2027.
Recent quarters exposed the strain. Intel missed some demand because it hadn’t reserved enough capacity for resurgent data-center chips. Its own product teams still send key designs to TSMC. Winning them back counts as priority one.
Naga Chandrasekaran runs Intel’s factory operations. He joined from Micron nearly two years ago. “Intel products alone, even in a wildly successful scenario, cannot fund the capital and filling the fabs and the scale that’s needed to be successful enough in a silicon business today,” he said. (The Next Web, May 8, 2026)
His candor stands out. Previous leadership under Pat Gelsinger often framed three years of losses and a 33 percent revenue drop from the 2021 peak in softer terms.
Culture adds another hurdle. Kevork Kechichian joined to lead the server chip unit after stints at Qualcomm and Arm. He encountered a habit of accepting delays. When teams slipped by weeks, he asked for recovery plans. They simply pushed the schedule out further. “I said, ‘What’s the recovery?’ and they came back with, the recovery is they adjusted the schedule to go another two weeks,” he recalled. (The Next Web, May 8, 2026)
Shifting that mindset consumes energy. Tan’s venture-capital background shows. He hires on high-level conversations more than detailed blueprints. He opens doors. He bets on people he trusts. That style succeeds in startups where failures stay contained. Semiconductor plants cost tens of billions. One missed process node can erase years of spending.
Recent shifts hint at adaptation. Tan now appears open to offering the 18A node to outside customers after initially reserving it mainly for internal use. Chief Financial Officer David Zinsner noted the change in March. The company also eyes advanced packaging deals with Google and Amazon that could reach billions annually. Those wins might arrive before full wafer revenue and deliver strong margins. (Reuters, March 4, 2026)
Still the core test holds. Can Intel produce chips that win market share and convince rivals to trust its fabs with their designs? Tan has two years, five years and ten years mapped out, he says. “Credibility comes from results.”
Investors bet on the relationships today. Factories must deliver tomorrow. The next twelve months will reveal whether Tan’s external triumphs translate into internal breakthroughs. As one fund manager observed, Tan received a difficult hand. Few others might play it better. But the table demands proof. (The Next Web, May 8, 2026)
WebProNews is an iEntry Publication