Former Intel CEO Pat Gelsinger has delivered a blistering diagnosis of the chipmaker’s long decline, telling the All-In podcast that the company that invented the microprocessor stopped being run by the people who understood it.
“Fundamentally this is a technology business and you need technologists running technology,” Gelsinger said. “One of the things that went off the rail was when it started to be run by business people as opposed to technical, the bean counters, the finance people.”
Gelsinger, a 34-year Intel veteran who returned as CEO in 2021 and was ousted in late 2024, painted a picture of a company that had systematically misallocated capital and abandoned the engineering culture built by Andy Grove, Gordon Moore, and Bob Noyce. Under the founders, roughly 15 of 20 executive staff members held PhDs. By the time Gelsinger walked back in, the company had not built a new factory in a decade.
The numbers tell the story. In the five to six years before Gelsinger’s return, Intel returned approximately US$100 billion (about £77 billion) to shareholders through dividends and buybacks. That capital, Gelsinger argued, should have funded new fabrication plants and next-generation extreme ultraviolet (EUV) lithography equipment, the tools Intel needed to maintain manufacturing parity with TSMC.
Instead, TSMC surged ahead. By 2021 the Taiwanese foundry was producing roughly five times Intel’s wafer output. By mid-2026, despite the Chips Act and Intel’s own foundry pivot under Gelsinger and his successor Lip-Bu Tan, the ratio had widened to roughly seven times.
Gelsinger pointed to specific strategic catastrophes he believes flowed from the same root cause. Intel passed on making chips for the iPhone, a decision that eventually led Apple to develop its own silicon, Apple Silicon, now a benchmark for the industry. And in 2009, one week after Gelsinger left the company, Intel killed its Larrabee GPU project, which aimed to turn x86 cores into a graphics-like processor.
“The world would have been so much different” had Larrabee survived, Gelsinger said. At the time, Nvidia’s CUDA platform was still a niche experiment. The project’s cancellation removed what could have been Intel’s vehicle into accelerated computing and AI, a market Nvidia now dominates with a market capitalization exceeding US$3 trillion.
Intel’s leadership during the CPU era “scoffed” at Nvidia’s machines, treating them as gaming hardware rather than a threat to the data center, a blind spot Gelsinger characterized as a failure of technical imagination compounded by financial engineering.
Gelsinger’s own turnaround effort between 2021 and 2024 focused on restoring Intel’s manufacturing competitiveness through a “five nodes in four years” roadmap and the creation of Intel Foundry Services, a bid to manufacture chips for external customers including would-be rivals. The company secured roughly US$8 billion in Chips Act grants and a US$3 billion Department of Defense contract. But the turnaround was incomplete when the board pushed him out in December 2024.
Since leaving Intel, Gelsinger has joined Playground Global as a venture partner and taken a role at Gloo, a faith-technology company. He continues to speak publicly about semiconductor strategy and US manufacturing competitiveness, warning that the Chips Act’s impact is real but its pace is insufficient: “We need to go faster, need to go more meaningful.”
Sources: Business Insider; All-In Podcast; Intel IR data on buybacks; Bloomberg on TSMC-Intel wafer output comparisons