Pat Gelsinger was one of the chip pioneer’s true believers. After his rescue strategy failed, the board lost confidence in him. How does this iconic American company survive?
When he was in vocational school, a teenage computer geek named Pat Gelsinger interviewed for a job that would define the trajectory of his entire life.
There were 12 candidates who had applied for an entry-level technician position based in Silicon Valley. He was the 12th interview. But when an engineering manager from one of America’s most innovative companies sized up Gelsinger, he was impressed by the farm boy from Pennsylvania who had never been on a plane before—and decided that he belonged at Intel.
“Smart, very aggressive and somewhat arrogant,” the interviewer wrote. “He’ll fit right in.”
Those were the characteristics that made Intel a goliath of the semiconductor industry.
As it turns out, they were also the qualities that have sparked an existential crisis for this iconic corporation that powered the growth of the American tech economy.
Gelsinger proved to be such a natural fit at the company that introduced the commercial microprocessor that he would spend the next three decades of his career there. Under the tutelage of legendary Chief Executive Andy Grove, Gelsinger climbed the ranks until he was named its first chief technology officer in 2000. He left the company in 2009 and was lured back in 2021 as CEO.
He returned at a time when the chip industry had never been so essential to geopolitics, society, national security and the entire global economy.
Chips are the engines of modern life. They are the indispensable pieces of technology behind our phones, computers, televisions and cars—and cyber espionage and advanced we-pons. In recent years, they have become the workhorses making artificial intelligence even smarter. And the pandemic-era chip shortage made it painfully apparent just how much we have come to rely on tiny, ridiculously intricate slabs of silicon.
All of which could have been quite lucrative for Intel. But the company had fallen behind in the race to make the best-performing chips with billions of microscopic transistors, ceding the cutting edge to Taiwan Semiconductor Manufacturing, or TSMC, and South Korea’s Samsung Electronics.
Gelsinger was hired with a clear mandate: Catch up.
The strategy that he articulated to restore Intel’s swagger ran utterly counter to the direction the industry had taken over the previous several decades. In that time, the chip business had evolved and basically split in two. Most chip companies have come to specialize in either designing chips (like Nvidia) or making chips (like TSMC). Because those businesses are radically different, only a handful of large chip companies still do both—like Intel.
Intel CEO Pat Gelsinger championed a comeback strategy that involved both designing and manufacturing chips.
But he thought he could steer this colossus into a new era of success by doubling down on manufacturing. That would require building new factories and moving at a breakneck speed to compete with nimble rivals that had surpassed the behemoth in chip-making technology.
The business of chips requires clairvoyance unlike almost any other industry, since CEOs have to make huge capital bets based on visions of the future that may or may not materialize.
From the start, Gelsinger acknowledged that his turnaround plan was bold and sought assurances from the board of directors that they would support his vision before agreeing to take the job. In a virtual meeting in January of 2021, he asked each board member to pledge their support for a strategy that was costly and ambitious but would represent one of the most sweeping turnarounds in American corporate history if it succeeded. They all signed on.
This week, he was abruptly pushed out after the board lost confidence in him, and the 63-year-old CEO was given the choice to retire or be removed, according to a person familiar with the matter.
His mission to save Intel ended with a sudden announcement on Monday morning that he was retiring and leaving the board.
When he was named CEO nearly four years ago, Nvidia and Intel had similar stock-market values. Since then, Nvidia gained $3 trillion and was crowned the most valuable company in the world, while Intel lost $150 billion and is no longer one of the 10 most valuable companies in the world of chips. Nvidia has been worth more than Apple and Microsoft; Intel is worth less than Boeing and Starbucks.
The company’s stock is down more than 60% since Gelsinger started as CEO, making Intel the single worst performer on the PHLX Semiconductor Index from his first day on the job until the last.
In late October, the company reported its largest quarterly loss ever, a stunning $16.6 billion net loss that was especially surprising because Wall Street analysts had forecast a loss of $1.1 billion. It has fallen so far that Intel was recently approached by Qualcomm with a takeover offer—once an unthinkable scenario.
An Intel spokeswoman said the company’s restructuring over the past few years had revitalized its chip technology “and laid the foundation for our future.” Efforts to be more efficient would create a leaner and more agile company to serve customers and create value for shareholders, she added.
But the company is in a fundamentally different position than it was even a few years ago, much less a few decades ago, when it was built around Grove’s famous mantra that only the paranoid survive.
So can Intel?
‘Make Intel Great Again’
Pat Gelsinger realized on his very first day at Intel in 1979 that he didn’t want to be a technician being told what to do. He wanted to be the engineer deciding what should be done.
A workaholic even when he was a teenager, he barely slept and logged so many hours that the company’s payroll department flagged his overtime. He was also a full-time student. Having moved across the country for a job with only his associate degree from Lincoln Technical Institute in Allentown, Pa., which he attended on a scholarship, Gelsinger took advantage of Intel’s generous tuition-reimbursement policies. The company covered his costs for a bachelor’s in electrical engineering from Santa Clara University and a master’s from Stanford University.
Gelsinger was 25 years old when he was put in charge of designing the industry’s most important chip, and he helped shepherd the next generations of semiconductors as the company became a household name with its classic “Intel Inside” marketing campaign.
When he was 40, he was named Intel’s first CTO. When he told his mother, she responded: “That’s great, honey. Now, what’s a CTO?”
Gelsinger spent decades at Intel, climbing the ranks until he was named chief technology officer in 2000.
A talented engineer and widely respected executive, Gelsinger kept hours that made other hard-charging corporate executives feel downright lazy, despite writing a book about work-life balance called “The Juggling Act” and developing a points system for himself so he wouldn’t stay at the office too late.
Grove’s departure in 1998 and Intel’s search for unexplored frontiers in the computing landscape presented challenges for the company—and Gelsinger.
By the 2000s, Intel had become the undisputed champion in central processors, the brains of personal computers, plus the servers that formed the backbone of the internet. But technology evolved and Intel couldn’t replicate that success when it tried to conquer other parts of the business. After he oversaw a scuttled project to compete with Nvidia in graphics chips, Gelsinger was ousted. He was devastated, but he landed at computing-infrastructure company EMC and then spent eight years as CEO at server-software giant VMware.
Meanwhile, Intel was struggling to keep up in advanced chip-making. As the costs of making the most cutting-edge chips ballooned, rival Advanced Micro Devices spun out its factory operations in 2009 to form GlobalFoundries—a sign that the future of the chip industry would be much different from its past.
Intel stood pat.
A series of manufacturing missteps clouded its long-term fortunes under CEO Brian Krzanich, who resigned in 2018 after having a relationship with an employee that violated company policy. Sales eroded further under Bob Swan, the former chief financial officer who took over as CEO.
The company’s failure to develop competitive mobile chips also meant that it largely missed out on the industry’s bo-m times in the 2010s.
Legendary Intel CEO Andy Grove was a mentor to Gelsinger.
Inside the company, people yearned for the glory days of Intel.
One executive even put a “MIGA” vanity plate on his Intel-blue Tesla, which employees interpreted as an acronym for Make Intel Great Again.
With the company under investor pressure, Intel board members approached Gelsinger about taking a board seat. That’s when he suggested an alternative path forward, one that defied industry trends and sounded like a bold move at what Grove might have called one of his “strategic inflection points.”
Gelsinger’s plan: Instead of adapting to a changed world with more outsourcing, Intel should go the other way and build more factories to become the king of American chip-making again.
And they bought his pitch—with one catch.
They didn’t just want a true believer like Gelsinger on the board. They wanted him back as CEO.
Bo-m—and bust
By the time he took over, the company was in such a dismal state that it wasn’t clear it could be saved by any CEO.
Gelsinger’s challenge was figuring out how to successfully implement plans that would have been more likely to succeed had they been launched years earlier.
Intel is one of the last remaining integrated device manufacturers, or IDMs, which means it both designs and manufactures chips. Most companies pick one or the other. Nvidia is often called a chip maker, but it doesn’t actually make chips. The physical act of etching circuits into silicon wafers takes place largely in the fabrication plants of TSMC, which pioneered the business of producing chips for clients—a model that changed the industry forever.
Gelsinger wanted to start a business like TSMC’s within Intel. To justify the huge capital outlays of building new chip facilities, which can cost $20 billion, Intel would make chips on contract for other companies. He hired a bevy of engineers and managers to handle the new business, recruiting heavily from the ranks of longtime Intel executives who had left after the company lost its mojo.
He was also the leading industry proponent of the 2022 Chips Act, which provided tens of billions of government dollars to fund chip factories and put American manufacturing back on the map. The government is now relying on Intel to be the only U.S.-based company capable of making advanced chips for the defense industry.
He called his strategy “IDM 2.0” and set an internal goal of making Intel the second-largest contract chip maker in the world behind TSMC by 2030.
But one problem was that Intel had no significant customers and was unlikely to peel them away from TSMC or Samsung without being able to show significantly better manufacturing capabilities.
There was also a culture problem. TSMC and other contract chip makers bend over backward to please customers and make sure they get exactly what they want. Intel wasn’t nearly as flexible. Its factories were used to making chips solely for one client: Intel.
Gelsinger, pictured with President Biden at the site of a planned Intel manufacturing facility, was the leading industry proponent of the 2022 Chips Act, which provided tens of billions of government dollars to fund chip factories.
To change the mindset inside Intel, Gelsinger looked to the outside. He discussed buying GlobalFoundries for about $30 billion before that potential deal fizzled. He then struck a deal for contract chip maker Tower Semiconductor, but it was torpedoed by Chinese regulators and Intel called it off.
The new CEO also promised to make five advancements in cutting-edge chip-making over four years. It was a blazing tempo within an industry where making just one leap can take years to pull off and typically involves many billions of dollars of engineering and manufacturing-equipment costs.
The pace of transformation was so critical that Gelsinger turned the word “Torrid” into a company rallying cry, printing it on T-shirts and even putting it on a sticker for his phone.
At first, it seemed like the winds might have shifted in Intel’s favor. During the pandemic, people working at home needed laptops and other equipment packed with Intel’s chips, as did the server farms that process internet traffic.
But that bump turned out to be a blip. Some senior executives grew skeptical of Intel’s strategy and left the company, and Gelsinger himself described the once-mighty company’s position as a “mud hole.”
Not even the AI bo-m could pull Intel out.
Intel had invested billions of dollars in AI chips long before the arrival of ChatGPT in late 2022, buying startups like Nervana Systems in 2016 and Habana Labs in 2019. But these days, most AI computation is handled by powerful graphics-processing units made almost exclusively by Nvidia.
Even as the demand for Nvidia’s chips remained insatiable, Gelsinger said in October that Intel’s AI chips would fall short of a revenue target of $500 million this year. At the exact moment he needed cash to make his turnaround plan work, tech’s richest companies were shifting their massive budgets toward Nvidia and other chip companies.
Now his exit could accelerate a potential breakup of Intel into separate manufacturing and design businesses—the very outcome he tried to avoid.
Gelsinger’s last day as CEO was Sunday. A deeply religious man, he often leaned on his faith to make business decisions and tweeted out a Bible verse every Sunday. That morning, he chose a psalm to mark the end of his time at the company that hired him decades ago, back when the future belonged to Intel.
“I will give thanks with my whole heart,” he posted. “I will tell of all your wonderful deeds.”