This article is based on a conversation between Lieven Vandersypen and Daniel Rodán Legrain as part of the Quantum Builders series, sponsored by Qblox. Watch the full webinar for more on spin qubits, the Delft ecosystem, and what it takes to scale semiconductor quantum processors.
Spin qubits had a quiet decade before they had a loud one.
For years, the platform sat behind superconducting qubits and trapped ions in the public conversation. Then at the APS Global Physics Summit earlier this year, the spin qubit session was packed, and the results from HRL, Intel, Diraq, the Tarucha group, and QuTech made it clear the field had shifted.
Lieven Vandersypen has been working on semiconductor qubits since 2001. He's a professor at TU Delft, director of research at QuTech, and one of the figures most associated with the spin qubit roadmap. The conversation traced how the field reached this moment, where it still falls short, and what the next five to ten years will require.
Vandersypen emerged from NMR quantum computing during his PhD, where he ran some of the earliest implementations of quantum algorithms on real hardware. That platform had clear scaling limits, and by 2001, he was looking for a successor.
Three things drew him to electron spins in quantum dots.
"If it were possible to build high-quality quantum bits based on semiconductors, based on the same technology that's driven the information revolution of past decades, that's a very powerful starting point," Vandersypen said.
The second was that the electron spin is the canonical two-level system, the version of a qubit that appears in physics textbooks. The third was personal.
"The first qubit had still to be built. You know, nothing had been done yet," he said. "If this field was going to go somewhere, which at the time I really had no idea, I would be there from the start."
He's clear-eyed about the uncertainty involved.
"For all I knew, it could also die within a few years, and nobody would talk about it anymore."
The early generations of spin qubit devices used gallium arsenide. The material was the workhorse of mesoscopic physics, but it carried a problem that nearly killed the field's prospects.
Every gallium and arsenic atom carries a nuclear spin. An electron confined in a quantum dot sits in contact with roughly a million of these nuclear spins, each pointing in a random direction. Statistical fluctuations in their orientation caused the electron spin to dephase within tens of nanoseconds. Coherence collapsed before useful operations could happen.
The path forward was known in principle. Build the quantum dots in silicon, germanium, or carbon, where nuclear spins are rare. The path forward in practice took most of a decade.
"It's been a painful road to fabricate devices that were sufficiently stable to be able to do, to perform qubit measurements," Vandersypen said. "By the time that you had tuned up all the gate voltages just right to start experiments, something in the background would shift, and you had to start over pretty much."
He was honest about how close he came to leaving the field.
"There was a time when I started thinking, what shall I do after I stop working on spin qubits because it no longer looks attractive."
The breakthrough came around 2013 and 2014, when teams at UNSW Sydney and elsewhere demonstrated stable single-spin Rabi oscillations in isotopically purified silicon. T2-star times jumped four orders of magnitude. Coherence stopped being the limiting factor.
"Seeing them work really boosted my motivation and confidence, and made me go all in," Vandersypen said.
The case for spin qubits comes down to compactness, manufacturability, and resilience.
A spin qubit is small. Vandersypen estimates roughly 10,000 of them would fit in the footprint of a single superconducting qubit. They can be patterned with lithography. The infrastructure of the semiconductor industry can be applied directly to their fabrication, their wiring, and their readout.
They also tolerate temperature in a way most quantum platforms don't.
"Even at, as we call it, very hot temperatures, as warm as a Kelvin, the qubits can be very well behaved," Vandersypen said. "There is a little bit of degradation in the performance compared to 10 millikelvin, but not too much."
That difference matters because cooling power at 1 Kelvin is orders of magnitude greater than at 10 millikelvin. It opens the door to running local classical electronics close to the qubit chip, which is part of why teams have invested in cryo-CMOS over the past decade.
He's careful not to oversell the platform.
"Every technology has its strengths and also its difficulties, and I also want to be straightforward about some of the challenges."
The biggest challenge is a direct consequence of the strength.
Spin qubits are small enough that getting wires to each of them is genuinely hard. Every qubit needs at least one control line, and at scale that creates a routing problem that lithography alone doesn't solve.
When asked where he'd spend a single magic wand on the field, his answer was direct.
"If I had a magic wand and I could solve the wiring challenge once and for all, really for good, scaling to arbitrary numbers of qubits, that's what I would pick," Vandersypen said. "Because then I think also the qubit personalities can be more easily dealt with."
The "personalities" line refers to a separate scaling problem. Microscopic variations between spin qubits mean each one behaves slightly differently and may need individually calibrated control signals.
"Personalities are wonderful in people. Everybody is a bit different. That's fantastic, that makes us rich as people," Vandersypen said. "In qubits, it's of course helpful if all qubits are really exactly alike. And that's not the case in these semiconductor qubits."
Cross-talk, by contrast, he treats as largely manageable. Capacitive cross-talk between neighboring quantum dots can be characterized as a sparse matrix and compensated for in software. The wiring problem is structural in a way that cross-talk isn't.
The APS March Meeting session on spin qubits was full, and Vandersypen credits the HRL team for some of the buzz. But he's clear the substance was real.
He identifies three categories of progress that converged.
The first was scaling at the chip level. Industrial teams at Intel, IMEC, HRL, and CEA-Leti have spent years applying semiconductor manufacturing methods to quantum chip fabrication. After a long buildup phase, the payoff is starting.
"Going from 10 qubits to 100 or a thousand or 10,000 at the level of the qubit plane is straightforward, because it's just repeating patterns which we can do with lithography," Vandersypen said.
The second was wiring. Several teams demonstrated multi-layer interconnect schemes capable of routing signals out of devices with hundreds to thousands of qubits, with corresponding progress on cryogenic control electronics and the superconducting cables that connect them to the qubit chip.
The third was conceptual. New architectural ideas, particularly around moving electrons across the chip rather than holding them in place, are reshaping how the field thinks about connectivity.
For the first 25 years of the platform, spin qubits were operated on while sitting in fixed quantum dots. Vandersypen's group is part of a wave demonstrating something different.
"Imagine that additionally you could take an electron and literally displace it some distance over the chip," he said. "What would that bring?"
The initial motivation was connecting separate qubit registers across a chip. But shuttling within a register turned out to bring something more useful: effectively all-to-all connectivity. Any qubit can be brought next to any other qubit in the array.
The deeper finding came when his group looked at what happens during the motion itself.
"You can also compute on the qubits while they are moving, and not only while they are moving, but you could operate on them controlled by the motion itself," Vandersypen said. "Depending on how fast or how far you displace an electron, you could perform one type of rotation on the qubit or another."
He cites parallel work from his QuTech colleague Menno Veldhorst on hopping gates, where the qubit moves between quantum dots with different quantization axes and the motion itself drives the rotation.
Vandersypen is convinced that shuttling will be part of any future spin qubit architecture.
"It relaxes constraints or demands for error correction. It relaxes hardware constraints. I think it's very powerful."
Vandersypen has been in Delft for 25 years and was scientific director of QuTech for four and a half. He's watched the ecosystem form in real time, and his read on what made it work is specific.
The first factor is reinvention. The department repositioned itself multiple times, from nanoscience to quantum nanoscience, then through the founding of QuTech in 2013 as a deliberate effort to expand beyond physics into computer architecture, circuit design, materials science, and error correction.
"Back in 2012, 2013, when somebody said ecosystem, it meant nothing to me. To me it felt like a buzzword which was empty and hollow," Vandersypen said. "Now I've experienced it."
QuTech has grown from around 60 people to over 300. The Delft-based quantum startups, taken together, surpassed QuTech in headcount about two years ago.
The second factor is collaboration as a design choice, not an accident. About ten years ago, Vandersypen ran an informal count of his joint publications within the quantum nanoscience department and found he'd co-authored papers with roughly 12 of the 18 PIs. The leadership of QuTech, he said, deliberately stimulated that kind of cross-group work.
The third factor, which comes up across his career, is industry collaboration. The partnership with Intel ran weekly meetings for six years. Devices, design feedback, and learning flowed both ways. The model has now expanded, with Intel, HRL, and IMEC sharing devices with multiple academic groups.
"There are only a few facilities in the world that can manufacture these devices," Vandersypen said. "To involve others in the feedback and in the design and in the next generations, I think, is very worthwhile."
Vandersypen has trained many of the people now leading spin qubit groups around the world. His advice for students is consistent across the conversation.
The hard skills matter. Some understanding of quantum physics and quantum information science is the baseline. Beyond that, he points to a meta-skill the field demands.
"Speak the language of fabrication experts, speak the language of software engineers, speak the language of circuit designers, and so on, and vice versa," he said. "Building a quantum computer is extremely hard. The quantum computer at scale is an extremely complex product system, and in this system, many aspects need to come together."
On choosing a specialization, his counsel runs against strategic optimization.
"Don't think too strategically, like, is this skill going to be needed five years from now or 10 years from now," Vandersypen said. "There will be room for good people regardless of what specific skills they have acquired. But following your passion, your interest, and being really good at something, becoming the world expert at something, I think that is a very valuable process."
Vandersypen expects commercial spin-based quantum processors to be in the field at scale within five to ten years. Whether they'll be solving meaningful computational problems by then is less certain, but they'll be tested, used, and stressed under real conditions.
He's confident that shuttling will be part of the architecture. He's also involved in a startup, Groove Quantum, built around germanium quantum dots, which he sees as a strong candidate platform.
The deeper shift he expects is cultural. The field has spent decades organized around physics groups led by individual PIs. The next phase requires systems engineering.
"A future quantum computer is not going to be built by a collection of physics PhD students led by a PI," Vandersypen said. "You need to think of this quantum computer as a system where everything needs to come together."
He listed the dependencies: materials choice, qubit encoding, control signal generation, timing precision, error correction scheme, connectivity, operating system, and applications. Each one constrains the others.
When asked what single breakthrough is still missing, he returned to the same answer.
"We have not yet demonstrated all of the necessary ideas for wiring up a million qubits. All the way to 10,000, I see it happening on a single chip. Beyond that, we have not yet demonstrated all the pieces that are needed to make that happen."
Three things stand out across the conversation.
The platform has crossed an inflection point. Material quality is no longer the binding constraint. Industrial fabrication is taking over from artisanal device-making, and the resulting chips are growing fast.
The hard problems are now architectural. Wiring, connectivity, and systems integration matter more than any single device improvement. Shuttling and similar ideas are reshaping what those architectures look like.
The next generation of progress won't come from any single team or any single discipline. It'll come from groups that can combine fabrication, control electronics, error correction, and software into a working system. That's the shift Vandersypen is betting Delft prepared the field for.
Scaling quantum systems requires more than individual advances. It requires coordination across hardware, software, and infrastructure. Qblox works alongside research teams, industry partners, and system builders to make that possible. If you're exploring how to move from experiments to systems, we'd be glad to connect.
Contact us to learn how Qblox supports scalable quantum systems.
