When Pat Gelsinger was asked after his initial departure from Intel in 2009 what differentiated his new employer EMC from Intel, he said: “Intel is more like a military organization.” This still holds some truth today: If the battle is lost, the general must go, and if a company’s numbers are not right, the CEO must go. This is the harsh corporate culture of publicly traded US companies.
However, the chip business is not that simple. While Intel recorded an adjusted net loss of two billion US dollars last quarter, Nvidia made a net profit of 16.6 billion. This success at Nvidia comes after years of investment in GPUs as computing accelerators, while most customers associated the company only with gaming chips. Today, Nvidia is the clear market leader in hardware for AI applications, a trend Intel missed due to a lack of investment. Now Intel is cutting costs significantly, and the issues are so severe that even a prominent figure like Gelsinger must leave.
What the company’s announcement regarding Gelsinger’s departure does not reveal is who will be his true successor. So far, there are only two “interim CEOs,” reminiscent of Apple in the 1990s. What about Intel’s commitment to its own chip manufacturing? Is a split like AMD’s in 2009 looming? All this remains unclear, and one can only guess what Intel’s board might have implied between the lines of its announcement.
No new strategy is apparent without a successor. It seems more like panic: get rid of him, regardless of what comes next. It’s like firing a football coach mid-season: the assistant continues, the team remains unchanged—costs must be saved—and usually, the championship isn’t won. Intel cannot afford to muddle through any longer.
Gelsinger’s strategy was clear: invest in semiconductor plants, not only for Intel’s products but also as a foundry for other chip developers. The model is TSMC. The need for semiconductors, not just for AI, is expected to keep growing in the coming years. Cars, smart homes, and almost everything that uses electricity are becoming more digital, requiring chips. The demand for semiconductors is so high that OpenAI CEO Sam Altman fantasized in the summer of 2024 about raising trillions of US dollars globally to build more chip factories. Intel risks missing another trend after smartphones, GPUs, and AI accelerators: flexible contract manufacturing. Intel has everything from chip design to manufacturing and packaging in-house.
If Intel wants to compete here, it must continue Gelsinger’s plan. The alternative is a much more painful split than AMD’s, as Intel’s current product range is much larger than its competitor’s fifteen years ago. This is part of the problem; Intel has spread itself too thin. To use the football metaphor one last time: Intel has retired the coach but hasn’t introduced a new team or a new playing philosophy. In 2017, colleague Martin Fischer wrote a commentary titled “The Survival of Intel is at Stake.” This is even more true today, especially after Gelsinger’s departure.