The $100B company trying to downsize its way back to relevance

Intel has a brutal plan to shrink its way into the future

You know the feeling. 

You’re stuck in a meeting that could’ve been an email, listening to someone explain a slide deck that’s already been shared twice.

The project has five owners, three timelines, and zero clear direction. 

It’s a familiar kind of corporate bloat - a slow, creeping inertia that eats away at innovation until all that’s left is process. For years, Intel embodied that exact dynamic.

Now, the $100B semiconductor giant is trying to shrink its way back into relevance.

Once an undisputed heavyweight, Intel is facing a very modern tech conundrum: it’s big, it’s slow, and it’s no longer dominating the market it helped define. Artificial intelligence has become the defining wave of this decade, and Intel has missed it. While Nvidia’s valuation soared past $2 trillion and its GPUs became the default engine for machine learning, Intel’s revenue in 2024 fell by $26 billion from its 2021 peak. That kind of drop isn’t just painful—it’s humiliating for a company that used to power nearly every PC on the planet.

Despite pulling in $12.7 billion in revenue in the first quarter of 2025 - slightly ahead of analyst expectations - the outlook for Q2 remained underwhelming. Intel forecasted revenue between $11.2 billion and $12.4 billion, a projection that left Wall Street disappointed and sent shares down more than 6% in after-hours trading. At a glance, it’s clear: Intel is still bringing in money, but the market has stopped seeing it as a growth story. 

To address that, Intel is going lean. Under new CEO Lip-Bu Tan, who took the reins in March, the company is undergoing a radical cultural reset - one that strips away the layers of management, meetings, and middle-of-the-road decisions that defined its recent years. It’s not subtle. Intel is cutting its workforce by 20%, enforcing a four-day return-to-office policy, delaying its $28 billion Ohio factory to 2030, and shaving $2 billion off its capital expenditure for 2025. The message is clear: get smaller, get sharper, and get back to building.

These changes aren’t happening in a vacuum. As of late 2024, Intel employed approximately 108,900 people, already down significantly from previous years. But under the latest restructuring plans, that number could shrink even further, with reports suggesting layoffs of more than 20% are on the table. Meanwhile, rivals like Nvidia operate with roughly one-third of Intel’s workforce and generate more than double the revenue. Taiwan Semiconductor Manufacturing Company (TSMC), another key competitor, has about 30,000 fewer employees and still outperforms Intel in nearly every metric. The data paints a stark picture: in today’s semiconductor landscape, size alone no longer signals strength.

Tan’s turnaround effort centres on cutting bureaucracy so that engineers can focus on engineering again. It’s the kind of corporate housecleaning that, while painful, could rekindle a culture of innovation that’s been buried under decades of risk-aversion. The internal message leans heavily on this theme - "rebuild the culture" - but it’s also clearly a bid to control costs in a year of geopolitical uncertainty and market volatility.

Intel’s recent quarterly bump was partially driven by customers stockpiling chips ahead of potential U.S - China trade restrictions. Those tensions are real. Beijing is reportedly considering tariffs as high as 85% on American-made semiconductors - a move that would put more than $8 billion of Intel’s annual exports at risk. It’s not a sustainable boost, just a defensive spike.

At the same time, Intel is still waiting on full disbursement from the CHIPS Act, a U.S. government programme designed to bring more semiconductor manufacturing back onshore. While Intel secured $1.1 billion in grants, it’s holding off on expanding capital expenditure until there's more certainty from Washington. Right now, the 2025 capex forecast remains in the $8 billion to $11 billion range.

But perhaps what makes Intel’s plan most interesting is what it isn’t doing. There’s no bold pivot into the metaverse. No buzzy AI acquisition. No desperate blockchain-adjacent rebrand. Instead, the company is opting for a strategy that feels almost minimalist: remove what isn’t working, protect what is, and try to reorient the business toward long-term product execution. In a world obsessed with fast pivots and hot narratives, Intel is taking a slower, more surgical approach.

Of course, the big question is whether that will be enough. Intel has missed multiple computing revolutions - mobile, graphics, and now AI. Its attempts to re-establish itself in manufacturing have been hit by delays and spiralling costs. And its competitors are not standing still. Nvidia is pushing further into AI infrastructure, and TSMC continues to lead in advanced chip production.

Yet, there’s a logic to Tan’s approach. Intel doesn’t need to win tomorrow’s market overnight. It just needs to survive long enough to catch the next big wave - whether that’s edge computing, advanced packaging, or something not even on the radar yet. By shrinking its overhead, clarifying its mission, and giving its engineers more autonomy, Intel might just buy itself the time it needs to catch up.

It’s an approach rooted in humility rather than hubris. 

In the end, Intel doesn’t want to be the biggest company in tech - it wants to be essential again. 

And sometimes, staying essential means knowing what to let go.