Intel (INTC) is in crisis. The company forced out CEO Pat Gelsinger on Monday, its stock price is down more than 50% on the year, and the grand plan to add third-party semiconductor manufacturing to its repertoire is suddenly more unclear than ever.
Gelsinger’s gambit to transform Intel into a US-based version of Taiwan’s TSMC that builds chips for its rivals is one of the reasons the company received $7.8 billion in CHIPS Act funding. But the foundry business, which it plans to operate as an independent subsidiary, has been hemorrhaging money and its biggest customer is still Intel itself.
Yes, the company has picked up deals to build chips for Microsoft (MSFT) and Amazon (AMZN), but that hasn’t quelled calls for Intel to kill its third-party foundry plan. Bank of America Global Research analyst Vivek Arya wrote in an investor note that Intel could go as far as fully separating its design and manufacturing businesses.
But ditching its fabs, or fabrication facilities, is a lot easier said than done, especially if Intel wants to hold on to those CHIPS Act funds.
“They just got the CHIPS Acts money. And if you look very closely at the details, it basically says that Intel cannot sell more than a controlling share of the manufacturing footprint,” explained Mario Morales, group vice president of enabling technologies and semiconductors at IDC. “So that means that Intel will have to maintain manufacturing control of their capacity.”
•USD
(INTC)
In other words, there’s no easy path forward for the US’s largest homegrown chip builder.
CHIPS Act funding is tied to Intel’s foundry
The Commerce Department and Intel announced the chipmaker would be getting CHIPS Act funding just last week, with the DoC saying the cash would help Intel expand its manufacturing capacity in the US, a key part of Gelsinger’s plan to help the company compete with TSMC and Samsung, two of the largest chipmakers in the world.
Gelsinger initially unveiled the ambitious agenda, called IDM 2.0, or integrated device manufacturing 2.0, in 2021, the same year he took the reins as CEO. At the time, he said Intel would create a world-class foundry business. But Gelsinger took over an Intel beset by problems.
Lack of investment and missed opportunities over the years meant the company failed to catch the mobile wave and fell behind TSMC in design capabilities. And building out a global manufacturing footprint isn’t cheap, which, for a company that was already hurting, wasn’t ideal.
Gelsinger had four years to make the turnaround happen, and while Intel’s chip design process is catching up, building chip manufacturing facilities takes time, more than Intel’s board seemed interested in giving Gelsinger.
“Some of the problems that Intel had, and still has, are not things you can solve overnight. They are multiyear issues,” said TECHnalysis founder and chief analyst Bob O’Donnell.
“Gelsinger started on a path to follow some of these things. And I think ultimately, at the highest level, what he chose to do, which is to get the manufacturing stuff back in shape and to work on design, is ultimately what the company needs to do.”
Now Intel has to make a decision: carry on with Gelsinger’s plan or cut back and focus on its own chip building capabilities or give up the business entirely. But doing that means saying goodbye to CHIPS Act money unless it can get approval from the Commerce Department.
Under the terms of the CHIPS Act agreement, Intel is prohibited from certain change of control transactions. If the company spins off its foundry business, it has to hold on to at least 50.1% of the voting rights of the entity if it’s a private company. If the new entity goes public, a third party can’t own 35% or more of it. That means Intel can’t simply cut its foundry business loose and go fabless if it wants to hold on to that CHIPS money.
But moving forward with the foundry model could be difficult.
“I think as a combined entity it's hard to survive, because the biggest problem is your potential customers for foundry are all your competitors against Intel products,” explained Gartner analyst Gaurav Gupta. “So how would they have trust in Intel foundry and become its customers, right?”
If Intel does end up pulling back on its third-party manufacturing ambitions, however, it could end up eating a good portion of the losses.
“You'd have to scale back. You'd have to lay off people. I'm sure there would be some sunk cost, but the alternative is you keep losing billions of dollars a quarter,” Citi head of US semiconductor research Chris Danely told Yahoo Finance. “So, I think you just need to scrap it. Take your lumps. If you have to, return the money to the government.”
While Danely believes Intel should drop its goal of serving as a global foundry for other companies, he says it should continue to build its own chips.
“I still think they should manufacture their own CPUs,” he explained. "There's been speculation for a long time that they're just going to turn the keys over to [TSMC] and have [TSMC] make everything. We don't think that that's the right move.”
Before that, however, Intel needs to find a new permanent CEO.
Email Daniel Howley at dhowley@yahoofinance.com. Follow him on Twitter at @DanielHowley.