Former Intel CEO Says Company Gave Shareholders $100 Billion It Should Have Spent on Chips

Pat Gelsinger, the former chief executive of Intel, was 18 when he went to work there, and he likes to joke that he went through puberty at the company. His mentors were the founders and engineers who built it, men like Andy Grove, Gordon Moore, and fellow chip pioneer Bob Noyce, leaders he describes as deeply technical.

He still remembers what that looked like up close. The first time he joined Intel's executive staff meeting, he says, about 15 of the 20 people in the room held doctorates.

That background is why Gelsinger says Intel went off course. In the five or six years before he returned as chief executive, he says, Intel handed about $100 billion to shareholders in dividends and stock buybacks. It was a staggering payout, and Gelsinger is blunt that he would rather have spent the money on technology.

He would have built, he says. Intel, he says, had passed on making chips for the iPhone. It had gone ten years without opening a new plant. And he asks, in his phrase, how Intel could not buy EUV machines, the costly tools that use extreme ultraviolet light to etch tiny circuits onto silicon.

The reason Intel did not make those bets, in his telling, was who was making the decisions. The company had come to be managed by finance specialists and executives, the bean counters, rather than technologists. On a spreadsheet, he argues, a new factory or a risky mobile chip looks like a lousy investment, unless you understand where the technology is heading.

A technologist, he says, spends when the economics alone look bad, because the technology makes it right. Intel stopped doing that, he says, and then watched Apple, Nvidia, and TSMC, the contract chipmaker that manufactures chips designed by others, pull ahead. When Gelsinger returned, no new plant had opened in ten years, the iPhone business had gone elsewhere, and that $100 billion had already been paid to shareholders.