The Startup Hammering Down Walls with Laser Precision

Inversion Semiconductor co-founder Rohan Karthik’s hands were covered with wires and sweat. The Mojave Desert sun was setting as he crouched, frantically connecting electronics. Rohan and his team had five days to assemble a rocket, and two hours left before their launch window closed.

"We cut it so close," he recalls. "It was 6 p.m. Pacific, which meant it was 2 a.m. in the UK, where the rest of the team was watching."

The five-meter rocket stood against the endless desert expanse, the culmination of 75 of Imperial College London's brightest minds. As he ducked into the concrete bunker and watched his creation tear into the sky, Rohan felt something shift. He'd tasted the impossible.

He just didn't know he'd soon be chasing something even harder.

"Rocket science is easier than what we're trying to do," Daniel Vega, Rohan’s co-founder, says. "We know rockets can get up into space. We’re working on a problem that no one has solved before."

What Inversion Semiconductor is building is a lithography machine, which uses light to etch the impossibly tiny circuits that power every piece of technology you touch. Your phone, your laptop, your car? They all depend on chips made by these machines.

Here's the problem: Only one company in the world makes the most advanced version, a Dutch firm that charges $400 million per unit. And the only way they create the precise wavelength of light needed is by exploding tiny drops of tin 50,000 times per second.

"We fire a high-powered laser into a gas," Rohan explains, "and that generates these really strong electric fields and a bubble-like structure that captures and accelerates electrons." The whole accelerator shrinks from kilometers to centimeters.

Two paths appeared before them: Apply their talents to the booming software industry, or aim for the impossible.

They knew the impossible path meant years of hammering at problems others had given up on. They chose it anyway.

Daniel's journey to the impossible began with a compass.

"My father gave it to me as a kid," he remembers. "It's a pretty magical device. It continues to point in the same direction no matter how many times you turn."

That childhood fascination led him to color-changing skin for satellites, then to building particle accelerators for cancer treatment. But it was in London, working with plasma accelerators, that the pieces clicked.

"I knew lithography used light to make patterns on chips," Daniel says. "And here was this brand new class of accelerators. There might be a way."

Meanwhile, Rohan had gone from leading rocket launches to designing chips that power 90 percent of the world's mobile phones for Arm. He'd seen the bottlenecks, the single points of failure stacked like dominoes.

When they met at Entrepreneurs First in London, Daniel pulled Rohan aside on day one.

"This is what I'm interested in. You have this knowledge. Would you want to do this together?"

The rest is history in the making.

Breaking through walls

Their initial outreach to Lawrence Berkeley National Lab went nowhere. But Rohan had learned something from his rocket days: If one door doesn't open, try another.

"If it doesn't work, you just weren't talking to the right person," he realized.

So he went one level up the org chart. Then another. When they finally broke through, they found scientists at Berkeley Lab who'd been waiting decades for someone to apply their compact accelerator research to something revolutionary.

"Smart people want to work on interesting problems," Rohan discovered. The talent pool for what they're building? Maybe a couple hundred people worldwide. But when they explained their vision, something remarkable happened. Leading experts from around the world started sending technical updates.

"There's merit in trying," Daniel reflects. "There's merit in doing something hard."

Every impossible leap starts the same way.

Whether it's a rocket in the desert or a machine that doens’t yet exist, the tools remain constant: Persistence, shamelessness and commitment when you're choosing between sleep and one more simulation.

"We operate with this one goal," Rohan says, borrowing from OpenAI's playbook. "Solve one impossible problem every year. If we can do that, building the lithography machine becomes inevitable."

To those standing at their own crossroads, choosing between the obvious and the impossible, Daniel offers this: "Nothing worth doing ever came easy. When you find yourself struggling, that's often a sign you're onto something important."

The difference between a dream and a breakthrough?

It's about which walls you choose to hammer down. And whether you're willing to keep swinging when everyone thinks you're crazy. The impossible only looks that way until someone does it.