“Let him believe that ...to his peril.”
Those pointed words from then Intel boss Paul Otellini back in 2006 were directed at rival AMD; specifically its executive vice-president of sales and marketing Henri Richard.
Back then, AMD was competing against Intel in the PC market, and it had Intel in its sights. Richard had said Intel had become complacent and focused on protecting its margins and, as a result, AMD had been making gains.
Otellini didn’t agree.
RM Block
But that was in 2006 when Intel was, if not at its peak, at least still riding the wave of the PC boom. A year later, the first iPhone was announced. It spelled the beginning of the slow death for the company in terms of its dominance of the chip sector.
Intel is still a significant player in the chip market but the PC and laptop sector has been in decline for some time. The market isn’t dead but it is no longer as significant as it once was.
While there was a brief surge of demand in the months following the outbreak of Covid as people scrambled to buy new equipment to work and learn from home, the surge was short-lived and sales have since fallen back to what has become more normal levels.
Partly because of its very success in that PC/laptop market, Intel has over those years seen its rivals overtake it. AMD is now the number two player in the burgeoning market for AI chips for data centres. Nvidia has soared in value on its dominance in the artificial intelligence sector, leaving Intel scrambling in its wake.
Last week, shares in Intel sank as much as 26 per cent after it announced a dramatic effort to right size and reinvent itself. It will trim costs and cut as much as 15,000 jobs in a bid to turn the business around. In a memo to staff, chief executive Pat Gelsinger laid out the path facing the company. “I have no illusions that the path in front of us will be easy,” he said. “You shouldn’t either.”
It is not the first time Intel has tried to slash costs; there have been several rounds of it in recent years as the company has tried to reorient the business.
Falling value
The business had been considered part of the dotcom era’s “four horsemen” alongside Dell, Microsoft and Cisco. In 2000, its stock market value peaked at almost $500 billion. It has never hit those lofty highs since, and its value now stands around $91 billion. That is less than 5 per cent of Nvidia and about 40 per cent of AMD.
While Intel dominated the market for PC chips, it gradually fell behind in chips for devices that required less power and therefore less expensive chips. Its Atom chip failed to set the industry alight.
Meanwhile, the server side of the business was hit by an increasing shift towards AI chips, with money being funnelled into the new technology.
“Intel has been one of the forgotten horsemen of technology the last couple decades – never overtaking its year 2000 highs and struggling to get earnings back to where they were before the AI revolution,” said Michael Schulman, chief investment officer of Running Point Capital.
The company that was founded on innovation had somehow lost its way, outflanked by rivals and floundering. But what happened to knock Intel so far from the forefront of the business it had once led?
One of its biggest miscalculations was with the iPhone. In an interview with The Atlantic following his departure from Intel, Otellini recalled how Intel had missed out on supplying Apple with the necessary chips to power the iPhone. Apple boss Steve Jobs had a price in mind; Intel wasn’t willing to drop that low. Jobs went elsewhere, and the rest is history.
At the time, the iPhone was an unknown quantity. Apple’s own sales predictions were modest in comparison to its eventual growth; one million handsets by the end of the year. For Intel, Otellini said, it wasn’t clear if the price cut would be outweighed by volumes. The company didn’t budge on price, and Apple walked away.
That decision would come back to haunt Otellini.
The iPhone ushered in a mobile revolution and Intel was caught off guard. It developed the Atom chip to try to turn the mobile tide in the company’s favour, offering Intel’s expertise in a low power chip that would power sophisticated smartphones.
Otellini said the target was the top 10 to 20 per cent of smartphones. But the gains failed to materialise and its efforts to capitalise on the smartphone chips design business ended in defeat in 2016.
Problems continued
Otellini had already departed the company by then, replaced by Brian Krzanich in 2013. But the problems continued.
Under Krzanich, Intel effectively stagnated, delaying the shift to smaller architecture while its rivals moved ahead. The 10 nanometre process was delayed and the company would also miss out on the AI chip revolution, ceding ground to Nvidia and AMD. In 2016, the business made some big cuts, losing 12,000 jobs. Industry watchers would later point to that as a key failure of his tenure, with Intel losing vital expertise as it tried to slash costs.
In the end, it wasn’t Krzanich’s performance as CEO that was his eventual downfall, but rather a workplace relationship.
His successor Bob Swan – the interim CEO who apparently never wanted the job – lasted only a couple of years before he was forced out by activist investors, unhappy with the company’s performance.
When Gelsinger, the current chief executive took over as head of Intel in 2021, the group was already foundering. But, after years of false starts, Gelsinger was seen as the right man for the right job: a technically minded executive with a solid business background, courtesy of his time at the helm of VMware.
The following year, in another blow, Apple moved away from Intel technology for its laptops, with the launch of its own M series chips. They would be manufactured by rival manufacturer TMSC.
But under Gelsinger, Intel had a plan to regain its mojo. He laid out an ambitious project: five generational leaps in four years that would put Intel back on top. That plan included $100 billion in spending in the US to build and expand factories, a significant investment programme in Europe that included a new plant in Germany and billions ploughed into upgrading facilities in Ireland.
It is one of the largest factory-building sprees in the industry’s history. Intel has called in some partners to help, with Brookfield Infrastructure Partners taking a 49 per cent stake in its semiconductor fabrication plant in Arizona, and Apollo taking a similar stake in a facility in Leixlip.
With that investment, Intel hopes to position itself as a force in the foundry world, manufacturing chips for other companies in the same way that TMSC and Samsung have. That is a big shift away from what Intel once excelled at: designing and building its own chips.
Intel would become another customer of its foundry, no more entitled to additional capacity than any of its other customers.
But Gelsinger is counting on it being the future of Intel.
The plan has a catch: Intel needs to persuade enough big companies to use its manufacturing services. That will take time. Meanwhile, it is ploughing billions into ramping up its production capacity. But at the end of last year, the foundry business was running at an operating loss of $7 billion.
Demand decline
Gelsinger had been hoping for a rapid rebound in semiconductor sales, previously predicting last year’s third quarter would be the nadir for the company’s performance. Instead, demand for Intel’s processors has continued to decline, and the outlook is gloomy.
Intel’s second-quarter results underscored its predicament. Sales are too weak, costs are too high and margins are too low. And the company’s finances are now too stretched to keep funding its expensive turnaround plan. Hence the need for the latest cuts.
“On one hand, Intel is witnessing the erosion of its traditional data centre business as chip buyers aggressively migrate towards AI chips. On the other hand, Intel is navigating a challenging and expensive transformation to a foundry model,” said Tejas Dessai, research analyst, Global X.
The company has been trying to push things, to move at a faster pace. Last September, Fab 34 in Leixlip began producing wafers using Intel 4 technology, which uses extreme ultraviolet (EUV) lithography. The process transfers patterns to a silicon wafer, creating the blueprints for integrated circuits.
The EUV system allows the company to print circuitry smaller and more precisely than before, and delivers significant improvements in performance and power efficiency. The first batch of processors, called Meteor Lake, shipped in January, with Intel confident it would usher in “a new generation of personal computers” with dedicated AI and machine learning capabilities.
But will it be enough? Or will the once mighty Intel be relegated to an also-ran in the tech industry?
The story holds a clear lesson for the tech industry, even for those who are currently riding high on the back of AI demand. No one is immune to the shifts of the marketplace. To borrow Otellini’s words: let them believe that to their peril.
As recently as January 2020, the company was worth more than AMD and Nvidia combined. Today, those competitors are collectively worth nearly $2.6 trillion, while Intel’s market value stands at just $95 billion, after shedding 57 per cent of its value since the start of the year. – Additional reporting: Reuters / The Financial Times Limited 2024
- Sign up for push alerts and have the best news, analysis and comment delivered directly to your phone
- Find The Irish Times on WhatsApp and stay up to date
- Our In The News podcast is now published daily – Find the latest episode here
