Intel's potential exit from advanced manufacturing- Impact on Oregon

Intel has been the driving force in semiconductor technology for nearly all its 57 years, setting the cadence for advances in computer technology that made the PC ubiquitous and the internet transformative.

For the last quarter-century, Intel has done that work at its Ronler Acres research campus in Hillsboro. Its Oregon scientists kept the company on the cutting edge with a succession of breakthroughs in transistor design, semiconductor materials and manufacturing technology.

Intel shocked Wall Street last week when the company said it may be done with all that.

Staying current on chip technology requires enormous spending — Intel’s capital budget is $18 billion this year — and the company said it isn’t selling enough of its own chips to pay the bills.

So if Intel can’t find a big outside client that wants to use Intel’s forthcoming 14A technology for its own chips, due in three or four years, the company said it might just give up.

“We face the prospect that it will not be economical to develop and manufacture Intel 14A and successor leading-edge nodes on a go-forward basis,” Intel wrote in a regulatory filing Thursday.

“Any decision to pause or discontinue our pursuit of Intel 14A and successor leading-edge process technologies,” it continued, “may be effectively irreversible.”

The news shocked and confused Wall Street, not least because Intel made the declaration deep in a securities filing rather than in a press release or analyst call. The company’s stock plunged 9% the next day as investors sorted through the immense ramifications. Intel declined to elaborate this week on Thursday’s filing.

However, Intel indicated it will continue designing chips and outsource advanced manufacturing to other chipmakers regardless of the fate of its advanced research. And Intel said it will continue making its chips with older technologies through at least 2030.

But the company admitted that shutting down its advanced manufacturing would be a risky strategy that might itself cost billions and leave Intel’s future in the hands of outsiders.

Oregon was already reeling as Intel slashed at least 5,400 local jobs over the past 11 months. The notion Intel might also wind down its advanced research puts its entire Oregon future at risk — and along with it, one of the state’s economic pillars.

Analysts wonder: What’s Intel’s vision?

Intel’s crisis dates back several years, to manufacturing setbacks that cost the company its technological leadership and left its chips years behind the state-of-the-art. That enabled rivals like AMD and ARM Holdings to take market share in PCs and data centers.

Meanwhile, Intel failed to develop its own technology to power advanced artificial intelligence.

Former CEO Pat Gelsinger planned to spend tens of billions to catch up by building new factories and making new chips, and by opening up Intel’s own factories to make semiconductors for other companies. The Biden administration promised $7.9 billion from the bipartisan CHIPS Act to help Intel along.

Intel made progress technologically, but not nearly enough. Annual sales fell by a third, to $53 billion, even as its spending soared. Losses ballooned and the company failed to attract even a single large, outside customer to use its factories.

So the board forced out Gelsinger and replaced him in March with a veteran semiconductor executive and former Intel board member, Lip-Bu Tan.

Intel was bloated, in Tan’s view, who disclosed plans this month to lay off 15,000 workers around the globe. He shut down a plant in Costa Rica, scrapped plans for factories in Germany and Poland, and delayed a planned Ohio expansion indefinitely.

“Unfortunately, the capacity investments we made over the last several years were well ahead of demand and were unwise and excessive,” Tan told investment analysts last week.

None of that surprised observers, given Intel’s anemic sales. But when Intel raised the idea Thursday that it might be done with advanced manufacturing, observers were left scratching their heads both by the declaration and the absence of a clear explanation from the company’s leadership.

“Intel has two things against it. One is the fact that, a) they’re laying people off; and, b) they don’t really project a positive vision for the company,” said Jim McGregor, a longtime semiconductor industry analyst with Tirias Research. “That’s something that we’re missing from Intel. We need that positive vision from Lip-Bu.”

After Intel fired Gelsinger last year, executives floated the idea of breaking up the company. The idea would be to split its chip design business from its manufacturing arm, called Intel Foundry.

Some investors had been pushing a breakup for years, predicting the two halves of the company would perform better independently. Others, though, doubted either half could stand on its own. And it was never clear how Intel could finance a breakup.

Just two months ago, Chief Financial Officer David Zinsner appeared to rule out a split when he told investors that Tan “isn’t thinking about massive changes.” And earlier this month, Intel promised Oregon would remain “the epicenter of our cutting-edge semiconductor research, technology development, and manufacturing.”

Intel undercut both those statements with Thursday’s regulatory filing.

Retreat comes with risks

“If we are unable to secure a significant external foundry customer for Intel 14A, our next generation semiconductor manufacturing process technology, we may pause or discontinue our pursuit of next generation leading-edge process technologies,” Intel said.

Intel may yet succeed in 14A. On Thursday, Tan said it is working with potential customers to custom design its new manufacturing process to suit them.

“That gives me a lot of more confidence that this time we have customers engaging early enough in the inception,” Tan said. “Customers are excited.”

And surely AMD, Apple, Nvidia and other big chip designers would like to have Intel as a credible alternative to their current contractor, industry leader Taiwan Semiconductor Manufacturing Co. Having another option could give chip designers more leverage in pricing and features.

Intel says it hopes to begin making chips with its 14A technology in 2028 or 2029. Given the ramp-up time it takes for new manufacturing technologies, Bernstein & Co. analyst Stacy Rasgon estimated that gives Intel no more than 18 months to “land a hero customer on 14A.”

If Intel doesn’t find a big client to rescue it, Rasgon said, the company risks being stuck outsourcing its advanced chips while making older, less profitable chips in its own factories.

“They might get the worst of both worlds,” Rasgon wrote in a note to investors.

Compounding the problem, he said, Intel may scare off potential foundry clients by acknowledging that it may not stay in the foundry business.

“We believe the disclosure itself may make it more difficult to attract major customers if they are not convinced of Intel’s commitment,” Rasgon said, “so we hope it is not self-fulfilling.”

Intel’s own filing laid out several other issues it will face if it shuts down advanced manufacturing:

• The company would be entirely dependent on outside manufacturers for advanced chips.
• It owns more than $100 billion in factories and equipment, whose utility and value would be severely diminished if the plants shut down or shift to making older, less profitable semiconductors.
• Private investors helped fund factories in Arizona and Ireland and Intel will have to pay them back if it doesn’t hit manufacturing targets.
• Top employees may leave, and Intel may not be able to fill key positions, if the company isn’t making the most advanced technology.

Intel is also waiting on $5.7 billion in pending CHIPS Act subsidies — including $850 million for which it has submitted claims that the Trump administration hasn’t paid. That money was supposed to help fund expansions to Intel’s U.S. factory network, including $1.9 billion in Oregon.

President Donald Trump has been slow to deliver on commitments the federal government made during the Biden administration. It’s possible, Rasgon said, that Intel is trying to apply pressure on Trump by raising the possibility that the country could lose the only leading-edge chipmaker based in the United States.

“One semi-plausible thesis around this is that it is a cry for help to the administration coupled with a veiled threat,” Rasgon and his colleagues wrote in a note Monday. He said the company might, implicitly, be telling the government: “Help us or we’ll blow it up.”

If this is gamesmanship, then Oregon is caught in the middle.

“The whole point of the CHIPS Act was to make sure the United States plays a leading-edge role in chips,” said Duncan Wyse, president of the Oregon Business Council. “And it’s hard to see how that can happen if you don’t have a leading-edge node in America, and that would be Ronler Acres.”

Intel relies on thousands of researchers and factory technicians at Ronler Acres — now formally known as Gordon Moore Park after the company’s co-founder — to develop each new manufacturing node. It’s the company’s largest and most sophisticated site anywhere in the world, bigger and more advanced than even Intel’s headquarters in Silicon Valley.

Intel’s Oregon workforce peaked in 2023, when the company had more than 23,000 people at Ronler Acres and its other campuses in Washington County. It cut 3,000 jobs last year and has laid off at least 2,400 more just this month, bringing Intel’s local headcount to its lowest point in more than a decade.

Still, Intel employs more Oregonians than any other business and the chip industry’s average wage — around $180,000 last year — is more than double the average across all professions. Thousands more contractors work to equip, supply and maintain its Hillsboro factories.

All of that work appears to be at risk if Intel stops making leading-edge chips.

Innovation has always defined Intel, according to McGregor, the Tirias analyst. And he said it’s no less important now than it has been in the past.

“I get the whole point that it’s really expensive to do the next process node, especially on your own, but I don’t see how they can be competitive without it,” McGregor said.

The semiconductor business is fiercely competitive, but McGregor said each company also builds on its rivals’ successes. And he said no company has contributed more to the industry than Intel.

While Intel is struggling, and Tan is struggling to communicate his vision for the business, McGregor said the company needs to keep developing and making new chips, and everyone needs Intel’s contribution.

“The whole industry fails,” he said, “if Intel does not succeed.”

-- Mike Rogoway covers Oregon technology and the state economy. Reach him at [email protected] or 503-294-7699.