Prediction: Intel Names a Major Outside Customer for Its Next Factory Process Before 2027

Intel (NASDAQ:INTC) stock has been on a huge run. Shares trade around $122 as of this writing, up about 20% in just the past week.

A big part of that run rests on the foundry (the business that makes chips for other companies). But the story is still missing its most important announcement: a major outside customer publicly committing to 14A, Intel's next manufacturing process. On the company's January earnings call, CEO Lip-Bu Tan said customers should begin making firm supplier decisions starting in the second half of this year, extending into the first half of 2027.

The front half of that window closes in about three months, and I think Intel makes it. My prediction is that before the year is out, the company -- or the customer itself -- puts a name on a major 14A commitment.

Image source: Intel.

The clock is Intel's own

On the same call, Tan said two potential customers were already working with 14A's early design kit and test chips.

Management added a spending rule, too. Intel won't build 14A capacity for outside customers until they commit. In other words, a signed name is what unlocks the next stage of the build-out.

By May, Tan was telling CNBC that Intel expected commitments from multiple foundry customers in the back half of 2026, while declining to name any of them.

From data to capacity

Chief financial officer David Zinsner told a Deutsche Bank conference in late August that 14A's defect rate is falling faster than Intel targeted, its best showing on a new process since 22nm, about 15 years ago.

Outside engagement has increased significantly, Zinsner said, and Tan's team is meeting prospective customers weekly.

And the conversations have moved past evaluating the technology's data. Customers are now asking how much capacity they can get. Intel, Zinsner said, has "conviction" it will land external 14A customers.

The company is spending like it, too. Intel raised about $23 billion in new capital this summer, money pointed in part at 14A infrastructure and long-lead equipment orders. Risk production of Intel's own products on 14A is slated for 2027, with high-volume manufacturing in 2028.

Sure, schedules slip and talks stall. But companies don't typically pre-order equipment and hold weekly customer meetings for deals they expect to fall through.

What counts as a name?

My prediction calls for a public announcement, from Intel or from the customer itself, naming a major company whose chips will be built on 14A. In July, Intel and Fortinet announced that Intel will build the cybersecurity specialist's next security processor -- on the older Intel 4 process -- making Fortinet the foundry's first named outside customer under Tan.

The Terafab project doesn't count, in my view. In April, Intel was tapped to supply 14A process technology to the chipmaking complex Elon Musk's companies are planning in Texas. The partnership could prove enormous, but plenty is still unsettled, including who will run the facilities. And a partner using Intel's technology at its own complex is different from a customer trusting Intel's own factories with its most advanced chips.

A name matters because the foundry's numbers can't make the case by themselves yet. Intel foundry's second-quarter operating loss narrowed to $2.1 billion from $3.2 billion a year earlier. Revenue grew 31% from a year ago, to $5.8 billion. But external customers accounted for just $293 million of that revenue, compared with $22 million a year before. Most of the increase reflects Altera becoming a customer after Intel deconsolidated it, not new foundry wins. A major named 14A customer would be the first hard evidence that outsiders will bet flagship products on Intel's newest process.

The stock already assumes a yes

I don't think a name would do much for the stock, though. Intel is priced at close to 60 times its expected earnings for next year. Investors, arguably, are already treating foundry wins as a matter of time.

What would prove my call wrong? The window runs through the first half of 2027, so decisions could land in the back half without anything breaking. A customer could also sign quietly -- after all, Tan declined to name clients for months before the Fortinet announcement. Or a prospect could stick with Taiwan Semiconductor Manufacturing, the market leader.

Still, management set the timeline in January, raised the money in August, and said customers were asking about capacity by September. I think the first name lands before New Year's.

My prediction, however, isn't a buy call. Even with a name in hand, shares would still cost about 60 times next year's earnings. I think the stock is just too expensive here.

Daniel Sparks and his clients do not have positions in any of the stocks mentioned. The Motley Fool has positions in and recommends Fortinet, Intel, and Taiwan Semiconductor Manufacturing. The Motley Fool has a disclosure policy.