Intel is collapsing. What went wrong?

For decades, the name Intel stood tall in the world of computing. It powered generations of personal computers, dominated the server market and pioneered breakthroughs in chip technology.

But today, the once-invincible tech giant finds itself staggering — its share price halved, AI ambitions faltering, factories idle, and tens of thousands of jobs being slashed. While its rivals soar, Intel is stuck grappling with missed opportunities, strategic missteps and a rapidly shifting technological landscape.

This is the story of how the mighty chipmaker lost its way — and whether it can find a path back.

A glorious past

Founded in 1968 by Robert Noyce and Gordon Moore, Intel's innovations helped transform California's Santa Clara Valley into the heart of Silicon Valley. It launched the world's first commercially produced microprocessor in 1971 and powered IBM's first personal computer a decade later.

Through the 1980s and 1990s, Intel reigned supreme, its Pentium chips and "Intel Inside" branding becoming household names. It was a founding member of the "Wintel" alliance with Microsoft and a major force in shaping the modern PC era.

By the early 2000s, Intel was among the world's most valuable companies. But cracks had already started to show.

Strategic missteps

Intel's troubles did not begin overnight. The seeds of decline were sown as early as 2007, when it declined to make chips for Apple's iPhone. Apple instead chose Arm-based chips, a decision that would later reshape the mobile industry. Intel also dismissed the significance of EUV (extreme ultraviolet lithography), which delayed investment in the very technology that would later power its competitors' dominance.

In the 2010s, delays in adopting next-gen manufacturing processes and a rigid attachment to its x86 architecture left Intel trailing behind rivals like AMD and TSMC. AMD, leaner and more agile, began eroding Intel's market share with its Zen-based processors. Nvidia, once focused on gaming GPUs, surged ahead in AI and data centre technologies.

Intel missed yet another wave — the mobile revolution — while Qualcomm and Apple raced ahead. Its inability to adapt, paired with growing bureaucratic inertia, left the company vulnerable as competitors evolved.

Gelsinger's return

When Pat Gelsinger returned as CEO in 2021, he was seen as a saviour. A 30-year Intel veteran, he came armed with big promises, chief among them a plan to regain manufacturing leadership through five new process nodes in four years, anchored by the ambitious '18A' technology, a novel method of delivering energy to chips and a new type of transistor.

Gelsinger invested billions into new fabs in Arizona, Ohio and Europe, hoping to transform Intel into a foundry powerhouse that could fight with TSMC.

But the promises began to falter. Intel offended its key manufacturing partner TSMC with dismissive remarks about Taiwan's geopolitical instability. The resulting fallout meant Intel lost access to discounted chip production, hurting margins. Meanwhile, the 18A process faced delays and underwhelming results — Broadcom's test wafers reportedly had yields below 20%, compared to TSMC's far superior performance.

The chip industry has moved on. Nvidia leads in AI. Arm powers the next generation of mobile and cloud systems. AMD continues to innovate and grow. Intel, once a leader, is now fighting to stay relevant.

Gelsinger also overpromised on Intel's AI ambitions. In 2023, he publicly claimed $1 billion in AI chip sales, even though internal forecasts were half that and supply had not been secured. Eventually, Intel had to scale back its expectations, damaging credibility.

AI, Arm and the new order

Intel's downfall has been strikingly visible in the AI boom. 

While Nvidia soared to a $3-trillion valuation by powering generative AI models like ChatGPT, Intel stumbled. It acquired multiple AI startups, but failed to integrate them or create a competitive product. Its Gaudi accelerator line, though promising, struggled to gain traction. Even AMD is now working with Nvidia to develop Arm-based AI chips, which signals a shift away from x86.

Arm, once dominant in mobile, is now expanding into desktops and servers. Apple's transition to Apple Silicon proved that Arm could deliver superior performance and energy efficiency. Qualcomm's Copilot PCs and Microsoft's exclusive reliance on Arm for AI PCs have only deepened Intel's irrelevance in the most dynamic segments of the market.

The move away from x86 is a technological shift that Intel has failed to keep up with.

Collapse and retrenchment

By late 2024, Intel's crisis had deepened. Its Q3 net loss reached nearly $17 billion, and its share price fell below $20. Gelsinger was quietly pushed out in December, and no successor was named. His exit was swift, and he did not even remain as an adviser.

Lip-Bu Tan, a respected venture capitalist and former board member, was brought in as the new CEO. But Tan quickly began dismantling many of Gelsinger's plans. He shut down Intel's automotive chip division, scrapped factory plans in Germany and Poland, and announced the company would no longer pursue cutting-edge chip manufacturing unless significant external demand emerged.

Intel has already laid off nearly 40,000 workers in two years and will shrink its workforce by another 10,000 by the end of 2025.

Can Intel recover?

Intel still retains deep technical talent, strong brand recognition, and manufacturing expertise. But the road ahead is long and uncertain. Its future rests on whether it can find customers for its upcoming 14A and 18A nodes, restore investor confidence, and offer a compelling product roadmap in AI and cloud computing.

The chip industry has moved on. Nvidia leads in AI. Arm powers the next generation of mobile and cloud systems. AMD continues to innovate and grow. Intel, once a leader, is now fighting to stay relevant.

For the US, the weight Intel carries is beyond making chips. The company symbolises national security, technological independence, and the enduring hope of rebuilding domestic industry. But that hope is hanging by a thread at this point.