THE ECONOMIST: Intel’s troubles deepen, as its boss makes an abrupt exit

When Pat Gelsinger took over as Intel’s chief executive in 2021 he seemed to possess the same impatient mindset as his mentor, Andy Grove, a former boss of the chipmaker famed for declaring that “Only the paranoid survive”. Barely a month into the job Mr Gelsinger unveiled a plan to restructure the business and advance through five generations of production technology within four years. Nearly four years on, however, it is Intel’s investors who have grown impatient. On December 2 Intel announced that Mr Gelsinger would be retiring. That his departure is effective immediately, with a permanent successor yet to be appointed, suggests it was hardly voluntary. It leaves both Intel and the incoming Trump administration in an awkward spot.

Mr Gelsinger’s ouster comes after the chipmaker posted a $US16.6 billion ($25.6b) loss in its quarterly results last month, the biggest in its history. In August the company suspended its dividend for the first time since 1992 and announced plans to fire more than 15 per cent of its roughly 130,000 employees. Since Mr Gelsinger took over as Intel’s boss its share price has plunged by three-fifths. By contrast the PHLX Semiconductor Index, an industry benchmark, has risen the same amount over that period.

In fairness, Mr Gelsinger was dealt a difficult hand. Intel had once been synonymous with cutting-edge semiconductor technology, and when Mr Gelsinger took over it still dominated the market for chips that run data centres and PCs. But it had missed the boom in smartphones and was ill-prepared for the one that would soon come in artificial intelligence (AI).

Following a series of technical setbacks, it had also fallen behind TSMC, a Taiwanese chipmaker, in the production of cutting-edge semiconductors. Mr Gelsinger’s response was to split the design and manufacturing sides of the business, allowing the former to choose the best foundries for its needs while freeing the latter to serve other chip designers. He declared, optimistically, that Intel’s new foundry business would become the world’s second-largest by 2030, behind only TSMC.

The strategy has fallen apart spectacularly. To turn Mr Gelsinger’s foundry vision into reality, Intel splurged $US100b on production sites across four American states — just as its profits were evaporating. Its core business of central processing units has slowed as AMD, a long-time rival, has nabbed customers, and it has failed to gain traction with its Gaudi AI chips.

To make up the funding shortfall, Intel has turned to private-equity firms as well as America’s government, which on November 26th announced that it would award the company nearly $US8b in subsidies under the chips Act. To make it easier to raise money from outside investors, the company said in September that it would turn its foundry business into an independent entity with its own board and financial statements, but stopped short of a full separation. Mr Gelsinger’s successor may have to be more radical. Qualcomm, a designer of smartphone chips, flirted with acquiring the company, but has reportedly abandoned the idea, deterred, perhaps, by Intel’s size and complexity.

The incoming administration of Donald Trump will have a role to play in the next stage of the saga. Vivek Ramaswamy, who will lead a new Department of Government Efficiency alongside Elon Musk, criticised the Biden administration’s decision to grant the $US8b subsidy to Intel in its waning days. Mr Trump has called the Chips Act “wasteful”.

At the same time, though, he has expressed his displeasure with Taiwan’s dominance over chip production and accused it of “stealing” America’s semiconductor industry. He now faces a choice between rescuing Intel from its mess, or relying on TSMC, which is building three advanced production facilities in America, to bring chipmaking home.