A year after being pushed out of Intel, Pat Gelsinger is still waking up at 4 a.m., still in the thick of the semiconductor wars — just on a different battlefield. Now a general partner at venture firm Playground Global, he’s working with 10 startups. But one portfolio company has captured an outsized share of his attention: xLight, a semiconductor startup that last Monday announced it has struck a preliminary deal for up to $150 million from the U.S. Commerce Department, with the government set to become a meaningful shareholder.
It’s a nice feather in the cap of Gelsinger, who spent 35 years across two stints at Intel before the board showed him the door late last year owing to a lack of confidence in his turnaround plans. But the xLight deal is also shining a spotlight on a trend that’s making people in Silicon Valley quietly uncomfortable: the Trump administration taking equity stakes in strategically important companies.
“What the hell happened to free enterprise?” California Governor Gavin Newsom asked at a speaking event this week, capturing the unease that’s rippling through an industry that has long prided itself on its free-market principles.
Speaking at one of TechCrunch’s StrictlyVC events at Playground Global, Gelsinger — who is xLight’s executive chairman — seemed unbothered by the philosophical debate. He’s more focused on his bet that xLight can solve what he sees as the semiconductor industry’s biggest bottleneck: lithography, the process of etching microscopic patterns onto silicon wafers. The startup is developing massive “free electron lasers” powered by particle accelerators that could revolutionize chip manufacturing. If the technology works at scale, that is.
“You know, I have this long-term mission to continue to see Moore’s Law in the semiconductor industry,” Gelsinger said, referencing the decades-old principle that computing power should double every two years. “We think this is the technology that will wake up Moore’s Law.”
The xLight deal is the first Chips and Science Act award under Trump’s second term, using funding earmarked for early-stage companies with promising technologies. Notably, the deal is currently at the letter of intent stage, meaning it’s not finalized and details could still change. When pressed on whether the funding could end up being double the announced amount — or potentially not materialize at all — Gelsinger was candid.
“We’ve agreed in principle on the terms, but like any of these contracts, there’s still work to get done,” he said.
Techcrunch event
San Francisco
|
October 13-15, 2026
The technology xLight is pursuing is pretty serious in both scale and ambition. The company plans to build machines roughly 100 meters by 50 meters — about the size of a football field — that will sit outside semiconductor fabrication plants. These free electron lasers would generate extreme ultraviolet light at wavelengths as precise as 2 nanometers, far more powerful than the 13.5 nanometer wavelengths currently used by ASML, the Dutch giant that utterly dominates the EUV lithography market.
“About half of the capital goes into lithography,” Gelsinger explained of the entire semiconductor industry. “In the middle of a lithography machine is light . . . [and] this ability to keep innovating for shorter wavelength, higher power light is the essence of being able to continue to innovate for more advanced semiconductors.
Leading xLight is Nicholas Kelez, whose background is unusual for the semiconductor world. Before founding xLight, Kelez led quantum computer development efforts at PsiQuantum (a Playground Global portfolio company) and spent two decades building large-scale X-ray science facilities at national labs including SLAC and Lawrence Berkeley, where he was chief engineer for the Linac Coherent Light Source.
So why is this viable now when ASML abandoned a similar approach almost a decade ago? “The difference was the technology wasn’t as mature,” explained Kelez, who was speaking at the event alongside Gelsinger. Back then, only a handful of extreme ultraviolet lithography (EUV) machines existed, and the industry had already sunk tens of billions into the incumbent technology. “It just wasn’t the time to take on something completely new and orthogonal.”
Now, with EUV ubiquitous in leading-edge semiconductor manufacturing and existing light source technology hitting its limits, the timing looks better. The key innovation, according to Kelez, is treating light like a utility rather than building it into each machine.
“We go away from building an integrated light source with the tool, which is what [ASML does] now and that fundamentally constrains you to make it smaller and less powerful,” he said. And instead, “We treat light the same way you treat electrical power or HVAC. We build outside the fab at utility scale and then distribute in.”
The company is aiming to produce its first silicon wafers by 2028 and have its first commercial system online by 2029.
There are, naturally, hurdles, though right now, competing with ASML directly does not appear to be one of them. In fact, the very opposite is true. “We’re working very closely with them to basically design how we integrate with an ASML scanner,” Kelez said. “So we’re working with both them, as well as their providers, [like] Zeiss, who does their optics.”
When asked whether Intel or other major chipmakers have committed to purchasing xLight’s technology, Gelsinger said they have not: “Nobody has committed yet, but the work is going on with everybody on the list that you would expect, and we’re having intense conversations with all of them.”
Meanwhile, the competitive landscape is heating up. In October, Substrate — a semiconductor manufacturing startup backed by Peter Thiel — announced it raised $100 million to develop U.S. chip fabs, including an EUV tool that sounds awfully similar to xLight’s approach. Gelsinger doesn’t see them as direct competition though.
“If Substrate is successful, they could be a customer for us,” he said, suggesting that Substrate is focused on building a full-stack lithography scanner that would ultimately need a free electron laser, which is exactly what xLight is developing.
Gelsinger’s relationship with the Trump administration adds another layer to the story. He brought up xLight to Commerce Secretary Howard Lutnick back in February, before Lutnick was confirmed. At that point, Kelez says, he’d already spent more than a year pitching xLight to the government as a way to bring chip manufacturing back to the U.S., but the new arrangement has drawn criticism from some who view the administration’s approach as overreach.
Gelsinger is unapologetic, framing it as necessary for national competitiveness. “I measure it by the results,” he said. “Does it drive the results that we want and that we need to reinvigorate our industrial policies? Many of our competitive countries don’t have such debates. They’re moving forward with the policies that are necessary to accomplish their competitive outcomes.”
He pointed to energy policy as another example. “How many nuclear reactors are being built in the U.S. today? Zero. How many being built in China today? Thirty-nine. Energy policy in a digital AI economy equals the economic capacity of the nation.”
For xLight, the government stake comes with minimal strings attached. The Commerce Department won’t have veto rights or a board seat, says Kelez (pictured above, right, with Gelsinger). No information rights, nothing, Gelsinger adds. “It’s a minority investment, in a non-governing way, but it also says we need this company to succeed for national interest.”
xLight has raised $40 million from investors including Playground Global and is planning another fundraising round next month, in January. Unlike fusion or quantum computing startups that need billions, Kelez said xLight’s path is more manageable. “This is not fusion or quantum,” he said. “We don’t need billions.”
The company also signed a letter of intent with New York to build its first machine at the New York CREATE site near Albany, though that agreement also needs finalization.
For Gelsinger, xLight is clearly more than just another portfolio company. It’s a chance to cement his relevance in the semiconductor industry that he helped build, even if his methods put him at odds with Silicon Valley’s traditional ethos.
Asked about navigating his principles in the current political environment, Gelsinger retreated to a more technocratic view of corporate leadership — one where the money is from the U.S. government, administrations are temporary, and CEOs must remain above the fray.
“CEOs and companies should neither be Republican or Democrat,” he said. “Your job is to accomplish the business objective, serve your investors, serve your shareholders. That is your objective. And as a result, you need to be able to figure out what policies are beneficial on the R side or what policies are beneficial in the D side, and be able to navigate through them.”
He added separately of that $150 million from the Trump administration, “Taxpayers will do well.”
When asked if working across 10 startups is enough for someone who used to run Intel, Gelsinger was emphatic: “Absolutely. The idea that I can now influence across such a wide range of technologies — I’m a deep tech guy at the core of who I am. My mind is so stretched here, and I’m just grateful that the Playground team would have me to join them and let me make them smarter and be a rookie venture capitalist.”
He paused, then added with a grin: “And I gave my wife back her weekends.”
It’s a nice thought, though anyone who knows Gelsinger’s reputation as a workaholic might wonder how long that arrangement will last.
