Swedish physicists cut quantum computing operation times by more than 1,000-fold

Physicists in Sweden have found a way to carry out a broad class of quantum computing operations more than 1,000 times faster than existing methods, addressing what researchers in the field have long described as one of the central obstacles to building a quantum computer that can run without being overwhelmed by errors.

The work, done by a team at Chalmers University of Technology in Gothenburg, does not build a new quantum computer or run a new calculation on one. It is a theoretical method, worked out mathematically and confirmed through simulation, for controlling delicate quantum states with far fewer steps than researchers had assumed were necessary. The paper appeared this summer in Physical Review Letters and has drawn attention from independent quantum-computing researchers and trade press in the weeks since, as the community assesses how directly the approach could be built into real hardware.

What the researchers did

The team, led by staff scientists Tangyou Huang and Lei Du with theoretical physicist Lingzhen Guo, worked on what are known as bosonic quantum codes, an approach to storing quantum information not in a single physical qubit but in patterns of microwave light inside a superconducting cavity. Bosonic codes are attractive because they can be built to resist certain kinds of errors automatically, but manipulating them has historically required a long sequence of small, carefully timed adjustments, sometimes thousands of repeated "driving" cycles, each one a fresh opportunity for outside noise to corrupt the state being built.

The new method, described in the paper "Single-Period Floquet Control of Bosonic Codes with Quantum Lattice Gates", collapses that long sequence into a single driving cycle. It does this using what the authors call quantum lattice gates, a set of pre-designed operations that Du has compared to prebuilt modules that snap together, rather than components that have to be assembled step by step from scratch. The full derivation and supporting calculations are posted on the preprint server arXiv, where the authors made the manuscript available ahead of formal publication.

The numbers behind the claim

The headline figure is a speedup of more than 1,000 times for a range of operations on bosonic quantum states, achieved by reducing the number of required driving periods from the thousands typically needed under earlier adiabatic protocols to one. That number matters because of a basic tradeoff in quantum hardware: qubits and bosonic quantum states are extraordinarily sensitive to their surroundings, and the longer an operation takes to run, the more exposure there is to the electrical noise, stray heat and even cosmic radiation that can knock a fragile quantum state off course. Cutting the operation time by three orders of magnitude correspondingly cuts the window in which such errors can accumulate.

That sensitivity is also why the result is aimed squarely at fault-tolerant quantum computing, the long-sought goal of building machines that can detect and correct their own errors faster than those errors pile up. Today's quantum processors, including bosonic-code devices, are still considered "noisy," meaning they can perform limited calculations before errors overwhelm the result. A widely cited estimate from the field holds that error rates need to fall by further orders of magnitude before quantum computers can reliably outperform classical machines on problems that matter, such as simulating new molecules or breaking current encryption schemes.

"Our method shows that a diverse range of quantum operations on bosonic states can be completed within a single driving cycle, rather than the several thousand cycles that have been required previously," said Lei Du, one of the paper's authors.

Context: a crowded race toward error correction

The result lands in a period of intense competition among approaches to quantum error correction, with companies and university labs around the world pursuing different physical platforms, from trapped ions to photonics to the superconducting circuits used at Chalmers. Bosonic codes occupy a particular niche in that landscape: rather than spreading one logical qubit's worth of protected information across many physical qubits, as some leading error-correction schemes do, they aim to encode redundancy directly into the physics of a single microwave cavity, potentially requiring less hardware overhead per protected qubit.

Chalmers has an institutional stake in seeing that approach succeed. The university coordinates the Wallenberg Centre for Quantum Technology, a national Swedish research program working toward a 100-qubit superconducting quantum computer, and the new control method is described by the researchers as a direct contribution to that broader effort. The theoretical work was supported by the Knut and Alice Wallenberg Foundation and China's National Natural Science Foundation, reflecting the collaboration between Huang and Du at Chalmers and Guo, who also holds an appointment at Tianjin University in China.

Who stands to benefit, and who is skeptical

The most immediate audience for the finding is other researchers working on bosonic quantum error correction, a smaller subfield within the broader push toward quantum computing that includes teams at other universities and at companies developing superconducting hardware. For them, a theoretical shortcut that reduces operation time by three orders of magnitude, if it holds up experimentally, would ease one of the practical constraints on building larger, more reliable bosonic-code processors.

Independent coverage of the paper, including a summary from the trade publication The Quantum Insider and a writeup carried by ScienceDaily summarizing the Chalmers announcement, has generally treated the result as a promising but early-stage theoretical advance rather than a demonstrated hardware breakthrough. That distinction matters: the paper's results come from mathematical analysis and numerical simulation of how quantum lattice gates would behave, not from running the protocol on a physical superconducting chip. Researchers in the field who were not involved in the study have not yet published a peer-reviewed experimental test of the method, and the authors themselves are careful to frame the work as a proposal awaiting laboratory verification.

What happens next

The authors say they are already in discussions with experimental groups at Chalmers about testing the method on real hardware, and they say they hope to see a demonstration "in the near future," though no date has been set. An experimental test would need to show that quantum lattice gates can be implemented with the fidelity the theory predicts once real-world imperfections in microwave control electronics and cavity fabrication are taken into account, a step that has tripped up other theoretically elegant quantum control schemes in the past.

If the approach survives that test, the next question is how directly it could be folded into the 100-qubit machine that the Wallenberg Centre for Quantum Technology is building toward by the end of the decade, and whether other groups working on bosonic codes elsewhere adopt the same single-cycle control strategy. For now, the result stands as a theoretical proof that a widely assumed limit on how quickly bosonic quantum states can be manipulated was not, in fact, a hard physical limit, only a limit of the control methods used until now.