New chip-supply deals and President Trump’s cheering are giving Intel momentum it hasn’t had in years. But the recent surge in the company’s stock belies years of technical missteps that investors shouldn’t be too quick to forget.
Excitement about Intel began to grow last August, when Trump took money committed to the company as part of a Biden-era US manufacturing push and converted it into equity. This gave the US government a 10% stake and Intel’s stock skyrocketed.
The stake also made it clear that doing business with Intel could be politically advantageous. Further announcements followed, including a promise by Nvidia to invest $5 billion and work with Intel on central processing unit design. Google joined the party in April with a collaboration on CPU and AI chips. Then Elon Musk turned to Intel for help with his mega-chip fab project. And Trump announced last week that Apple had agreed to work with Intel on designing and making chips.
All this commotion has sent Intel’s stock up more than 550% in the past year, giving it a market valuation of more than $700 billion. This is a huge deal for a company that has been mired for years in what former Chief Executive Pat Gelsinger once called a “mud pit.”
Intel has made some progress climbing out of that hole under CEO Lip-Boo Tan, who took the helm last year. But it’s not clear what’s behind the technical struggles that have plagued the company over the past decade.
In recent months Intel has begun shipping chips made using its state-of-the-art “18A” manufacturing process, development of which began under Gelsinger. This was a major milestone for Intel. That would help Intel compete more closely with contract chip giant Taiwan Semiconductor Manufacturing Co and its customers, which include Nvidia and Intel rival Advanced Micro Devices.
However, it remains an unresolved question whether Intel can make its advanced chips financially worthwhile.
Chip makers typically start making new generations of chips with much smaller transistors by testing them in laboratories, then move production to factories where they can be made in much larger quantities.
At the same time, manufacturers strive to improve their yields, or the number of working chips they can get from a silicon wafer. If they can’t be manufactured in high volumes with good yields – how good depends on the size of the chips and how advanced the manufacturing process is – then it makes less financial sense to produce them.
And that’s the pickle that Intel is in. The company won’t say what its yields are for its 18A chips, but it’s clear they’re not a financial home run yet.
Chief Financial Officer David Zinsner said at an analyst conference this month that yields are “not up to a level that they are at least neutral to, if not, expected to drive overall company gross margin growth over time.” In other words, they are not helping Intel financially yet.
It is also not clear whether Intel’s agreement with Nvidia, Google or Apple will prove to be commercially dangerous. If Intel can’t make the chips that customers want at a financially meaningful profit, it’s hard to imagine that those deals will yield much sustained new revenue or profit.
Working in Intel’s favor is that many of TSMC’s customers want alternative suppliers. Only Intel and Samsung Electronics have a real chance to play that role. But the world’s biggest chip designers have been quietly exploring using Intel’s factories for years without any concrete results. The Intel division that houses the factories that make most of the chips that Intel designs itself, reported revenue of about $5.4 billion in the first quarter, with an operating loss of $2.4 billion.
Intel executives say they are making good progress in manufacturing. The company said last week that a better-performing version of the 18A manufacturing process has entered risk production, meaning it is one step closer to high-volume manufacturing. Officials also say the next generation 14A process is on track.
It’s tempting to believe them all. Bank of America analyst Vivek Arya double-upgraded Intel from “underperform” to “buy” earlier this month, on the basis that it was attracting new outside customers, advancing its chipmaking technology and benefiting heavily from the shift toward digital agents within AI. And in fact, AI agents use more CPU computing power, which plays into Intel’s strength as a major supplier of those chips.
But Intel’s record of broken promises is also impossible to ignore. In 2015, then-Chief Executive Brian Krzanich said that Intel would produce 10-nanometer chips within the next two years – the most futuristic in the world.
Those chips didn’t reach high volumes until 2019, allowing TSMC to gain a lead over Intel in cutting-edge manufacturing — a lead it still holds. Intel’s subsequent 7-nanometer chips were also delayed, adding to the company’s troubles.
These days, investing in Intel is a game of speed. The hope is that Trump will help boost business for a company that is increasingly taking on the role of America’s national chip-manufacturing champion. Whether Intel can expand its manufacturing capacity enough to pay for those orders is another matter.
Write to Asa Fitch at asa.fitch@wsj.com
