There is a cruel arithmetic at the heart of quantum computing. Every operation takes time, and every moment of time is an opportunity for the universe to whisper noise into your qubit. Electrical flickers. A passing cosmic ray. The thermal jitter of atoms that refuse to sit still. The longer your gate takes, the more likely your delicate quantum state drifts into something you never asked for.
For years, the field has pursued a strategy that sounds almost paradoxical: fight slowness with more slowness. To create the exotic quantum states needed for error-correcting bosonic codes, researchers would guide their systems through thousands of repeated driving cycles, building the state brick by brick, pulse by pulse. Each cycle bought a little more fidelity — and introduced a little more risk.
In September 2026, a team at Chalmers University of Technology in Sweden proposed a different path. Rather than thousands of cycles, their method completes a broad class of quantum operations in one. A thousandfold shortcut, published in Physical Review Letters, that could reshape how we think about speed and error in the race toward fault-tolerant quantum computing.[1]
Bosonic Codes: Storing Secrets in Light
Most quantum computers store information in individual qubits — two-level systems, like the spin of an electron or the energy state of a superconducting island. Bosonic codes take a different approach. They encode quantum information not in a particle, but in a field: the microwave resonances that ring inside superconducting circuits, the standing waves of light trapped in tiny cavities.
The advantage is intrinsic protection. Because the information is spread across many photons in a continuous variable, certain types of errors — photon loss, for instance — can be detected and corrected without destroying the encoded state. The qubit becomes a pattern in the field, not a property of a single object.
The trade-off is control. To manipulate that pattern, you need to nudge the field into precise quantum states. And that nudging, in the conventional approach, has been painstakingly slow.
The Floquet Trap
The standard technique is called Floquet control: you apply a periodic driving signal, over and over, and each cycle inches the system closer to the target state. It works. It is well understood. But it is also a recipe for accumulated vulnerability.
"The fundamental building blocks of quantum computers, known as qubits, are so sensitive that even the smallest disturbance can cause the quantum state to deviate from the target," says Lei Du, a researcher at Chalmers and lead author of the study. "If too many errors accumulate before they can be corrected, the computation can fail."[1]
Thousands of cycles means thousands of chances for decoherence. The very process meant to protect quantum information ends up exposing it, slowly, across the duration of the gate.
Quantum Lattice Gates: The Pre-Built Module
Du, together with Tangyou Huang and Lingzhen Guo, found a way out. Their solution rests on quantum lattice gates — a universal set of operations they had developed earlier, now deployed in a radically more efficient way.
The insight is geometric. Rather than constructing a target state step by step, the quantum lattice gate acts like a pre-fabricated module. You design the gate once, and a single driving cycle snaps the system into place. No iteration. No accumulation. Just one pulse, one period, one transformation.
"You can think of it like building a large Lego castle," says Huang. "Instead of assembling it brick by brick and risking mistakes along the way, quantum lattice gates act like pre-built Lego modules that can be connected quickly and efficiently."[1]
The result: operations that previously required thousands of Floquet cycles now complete in a single period. More than a thousand times faster. And because the gate is over almost before the environment has time to react, the error budget shrinks accordingly.
Built for the Hardware We Already Have
Perhaps the most appealing feature of the Chalmers approach is its compatibility. The method is designed for existing superconducting quantum circuit platforms — the same technology that Google, IBM, and Rigetti have spent years refining. No exotic new hardware required. No cryogenic revolution. Just a smarter way to drive the circuits we already know how to make.
Chalmers itself is betting on this technology. The university is currently developing a 100-qubit superconducting quantum computer as part of the Wallenberg Centre for Quantum Technology (WACQT). The researchers are already in discussions with colleagues about experimental demonstrations.
"A key advantage of our approach is that it can be implemented using existing superconducting quantum circuit platforms," says Huang. "We hope to see a demonstration of the method in the near future."[1]
What This Means for the Fault-Tolerance Horizon
Quantum error correction is often discussed in terms of thresholds — the error rate below which a logical qubit can be protected indefinitely, provided you have enough physical qubits and fast enough operations. But speed matters just as much as fidelity. A surface code, for instance, requires constant rounds of syndrome measurement and correction. If each round takes too long, errors slip through the gaps between corrections.
The Chalmers result attacks this problem from the gate level. By making bosonic-code operations fast enough that decoherence barely has time to act, it opens the door to using these intrinsically protected codes in real-time error-correction protocols. The dream is not just a logical qubit that lasts longer, but a logical qubit that can be manipulated on timescales competitive with the errors themselves.
We are not there yet. The work is theoretical and numerical, and experimental validation will reveal where the real limits lie. But the direction is clear. The field has spent years making qubits better. Now it is learning to make the operations on those qubits faster — and sometimes, speed is the best form of protection.
The Blink and the Horizon
There is something almost cinematic about it. A gate that, from the outside, looks like a blink. Inside that blink, a thousandfold compression of what used to be a slow, fragile ritual. The quantum state does not know it was supposed to take longer. It simply arrives, transformed, before the noise has time to notice.
Fault-tolerant quantum computing remains a hard problem. It will require advances in materials, fabrication, control electronics, and algorithm design. No single paper solves it. But the Chalmers result removes a bottleneck that many assumed was structural — the seemingly unavoidable cost of building protected quantum states cycle by cycle.
Sometimes the way forward is not to build better walls against noise, but to move so fast that noise never catches you.
References
- T. Huang, L. Du, and L. Guo, "Single-Period Floquet Control of Bosonic Codes with Quantum Lattice Gates," Physical Review Letters 137, 060501 (2026). DOI: 10.1103/PhysRevLett.137.060501 [arXiv:2505.14404]
- Chalmers University of Technology, "1,000 times faster operations bring reliable quantum computing a step closer," News Release, September 10, 2026. chalmers.se
- ScienceDaily, "Scientists just made quantum computer operations 1,000 times faster," September 11, 2026. sciencedaily.com