A new quantum-computing experiment points to a simple but important idea: future quantum chips may work better if some qubits can move.
The Ars Technica article covers research from QuTech, TU Delft and TNO, published in Nature, showing that electron-spin qubits in silicon can be transported across a chip, brought together for a two-qubit operation, and then used in a short-distance quantum state teleportation demonstration. The result does not mean practical quantum computers are ready, but it addresses a real design problem in scaling quantum hardware.
What happened
Researchers led by Lieven Vandersypen at QuTech and Delft University of Technology reported a silicon spin-qubit experiment in Nature on 6 May 2026. The paper is titled “Two-qubit logic and teleportation with mobile spin qubits in silicon.”
The team used electron spins as qubits. In this type of device, each electron is held inside a tiny electrically defined trap called a quantum dot. Instead of keeping the qubits fixed in place, the researchers used moving electrical fields to shuttle the electrons across the chip in travelling potential wells, often described as a nanoscale conveyor-belt approach.
The key step was not only moving the qubits, but making them interact while in motion. The researchers moved two electron-spin qubits toward each other from separate positions, controlled how closely they approached, and used the resulting exchange interaction to perform a two-qubit operation.
According to the Nature paper, the experiment achieved an average two-qubit gate fidelity of about 99 percent when the two spins were each shuttled by 120 nanometres, for a total displacement of 240 nanometres. The team also reported conditional post-selected quantum state teleportation between qubits separated by 320 nanometres, with an average gate fidelity of 87 percent.
The bigger picture
Quantum computers need many qubits that can interact reliably. The hard part is not only making good qubits, but also connecting them in useful ways without creating an impossible wiring and control problem.
Many solid-state qubit designs, including semiconductor spin qubits, are attractive because they may benefit from semiconductor manufacturing methods. But fixed qubits often interact most naturally with close neighbours. That can make routing information across a larger processor difficult.
Mobile qubits could help by allowing quantum information to be moved to dedicated interaction, readout or entanglement zones. Instead of building a separate high-quality connection for every possible pair of qubits, a chip could potentially move qubits to where they need to meet.
This idea is already familiar in some atomic quantum systems, such as trapped ions and neutral atoms, where particles can be rearranged or transported. The new work matters because it shows a version of that flexibility in a silicon semiconductor platform.
Why this may be happening
The research is driven by a scaling problem. Small quantum processors can demonstrate useful physics, but large fault-tolerant quantum computers will require many physical qubits, high-fidelity operations, and enough connectivity to support error correction.
Mobile spin qubits offer one possible path around the connectivity bottleneck. If qubits can be shuttled between storage zones, operation zones and readout zones, future chips may need fewer fixed point-to-point couplings and less crowded control wiring.
The Nature paper also argues that mobile qubits could let processors adapt their connectivity patterns during operation and use different layouts for different quantum error-correction schemes. That is important because error correction is expected to be one of the biggest overheads in practical quantum computing.
The QuTech explanation adds that the motion itself can be used as part of the control method. In simple terms, distance and timing become tools for making two qubits interact.
What it means
For quantum hardware, the result is a proof-of-principle step toward more reconfigurable semiconductor processors. It suggests that silicon spin qubits may not have to remain locked into rigid nearest-neighbour layouts.
For chip design, the work points to a possible architecture with storage areas, conveyor-like transport channels, shared interaction zones and fixed readout stations. That kind of layout could reduce routing complexity if it can be scaled.
For the semiconductor industry, the result is interesting because silicon-based qubits may be more compatible with existing chipmaking ideas than some other platforms. That does not mean today’s chip factories can immediately produce full quantum processors, but it strengthens the case for semiconductor-based quantum hardware as a long-term route.
For the public, the main takeaway is more modest than the hype around quantum computing often suggests. This is not a general-purpose quantum computer. It is a laboratory demonstration showing that mobile spin qubits can perform meaningful operations in silicon.
What is still unclear
- It is not yet clear how well the approach will scale from a small linear device to large two-dimensional quantum processors.
- The 87 percent teleportation fidelity is useful for a demonstration, but future fault-tolerant systems will need much stronger overall performance across many operations.
- Large systems would still need better integration of control electronics, cooling, wiring, calibration and error correction.
- Manufacturing compatibility does not automatically mean high yield, low cost or commercial readiness.
- The best architecture for future semiconductor quantum computers remains unsettled.
Final takeaway
The research shows that silicon spin qubits can be more flexible than a fixed-dot layout suggests. By moving electron-spin qubits on a chip, researchers demonstrated a way to bring distant qubits together, make them interact, and transfer quantum information across a short distance.
The important point is not that practical quantum computers are suddenly near. The important point is that one of the hardest parts of quantum-chip design—how to connect many qubits without overwhelming the hardware—may have a new possible answer.
Sources / references
- Ars Technica: Manufacturing qubits that can move.
- Nature: Two-qubit logic and teleportation with mobile spin qubits in silicon.
- QuTech: The qubit: there and back again.
- TNO: The qubit: there and back again.
- Nature News & Views: Nanoscale “conveyor belt” teleports quantum state of electron.
- Phys.org: Mobile qubits on a chip move us a step closer to everyday quantum computers.
- arXiv: Two-qubit logic and teleportation with mobile spin qubits in silicon.
