“By drastically reducing the chance of error, this work significantly reduces the infrastructure required for error correction, opening the way for future quantum computers to be smaller, faster, and more efficient,” said University of Oxford graduate student in physics Molly Smith and co-senior author on a study that has established a new international standard for quantum logic fidelity. For researchers in quantum computing, the statement is more than a milestone it indicates a sea change in how the next generation of quantum machines could be constructed.
The recent demonstration by Oxford’s team of a single-qubit gate error rate of just 0.000015% or one error per 6.7 million operations is the lowest error achieved for any quantum computing platform to date. This jump, almost an order of magnitude improvement over their own best prior result, was made utilizing a single trapped calcium-43 ion controlled not by the usual laser beams but by carefully tuned microwave pulses supplied via a chip-based resonator. The experiment was performed at room temperature and without magnetic shielding in a surface-electrode trap microfabricated a design that not only makes the hardware simpler but also increases stability and scalability Physicists in the University of Oxford have established a new world record for quantum logic fidelity.
The key to this unprecedented fidelity is a combination of technical innovations. To start with, the stability of a hyperfine “atomic clock” transition of the calcium ion, which is known to possess a long coherence time of a record 70 seconds, was the solid basis. Second, the team used automated amplitude and frequency drift correction, enabling the system to keep itself in top shape and keeping control errors down below 10⁻⁸. This automated error loop, coupled with stringent randomized benchmarking where blocks of random Clifford gates were implemented and the state fidelity decay was examined validated that the measured error rate was no statistical anomaly but a real physical limitation of the system The Oxford experiment utilized a single trapped ion as a qubit, taking advantage of a hyperfine “clock” transition.
To get an idea of how great a feat this is, note that the failure rate is so low that an individual stands statistically more chance of being hit by lightning in a single year than for one of these quantum gates to break. “As far as we are aware, this is the most accurate qubit operation ever recorded anywhere in the world,” said Professor David Lucas, co-author on the paper. The implications are profound. In quantum computing, every operation be it a flip, rotation, or entanglement carries the risk of error. When millions of such operations are chained together in a complex algorithm, even a modest error rate can render the final result useless. This is why quantum error correction, a technique that encodes logical qubits on numerous physical qubits to detect and correct errors, is a foundation of the field Quantum error correction is conjectured to implement high-fidelity logical qubits.
But error correction has a high cost: the lower the physical error rate, the more redundant qubits and operations required for reliability. The record-breaking fidelity of the Oxford team implies that error correction overhead can be significantly cut back, and the size, cost, and complexity of next-generation quantum computers reduced. As one source describes, minimizing the error rate of basic qubit operations is a major challenge that directly affects the scale, cost, and performance of quantum computers The result of this research shows a possibility of significantly reducing the number of qubits required for quantum error correction.
This outcome also highlights an increasing divergence between quantum hardware approaches. Superconducting qubits, which are led by industry leaders such as IBM and Google, provide fast gate speeds and scaling through easier semiconductor fabrication, but need cryogenic cooling and are still more prone to decoherence and noise. Trapped-ion platforms, on the other hand, have for a long time been praised exceptional coherence times and low error rates, but have slower gate speeds and are hard to scale. The Oxford experiment, using microwave-controlled gates rather than lasers, shows a way towards less expensive, more resilient, and easier-to-integrate control systems perhaps shifting the scales even further in favor of ion traps for high-fidelity use Ion traps are known for their long coherence times, which means quantum information can be stored in ion qubits for longer periods without degradation.
But as the Oxford group and others are quick to point out, the breakthrough is not a silver bullet. As single-qubit gates approach perfection, two-qubit (entangling) gates remain the bottleneck, with error rates still at 1 in 2,000 several orders of magnitude above the new single-qubit benchmark. The journey to wholly fault-tolerant, utility-quantum computers will need comparable improvements in multi-qubit operations as well as advances in SPAM fidelity, which is several orders behind gate fidelity at the moment and two-qubit gates still have significantly higher error rates – around 1 in 2,000 in the best demonstrations to date.
Nevertheless, the Oxford result sets a solid platform for the future of quantum engineering. As fault-tolerant architectures and error-correction codes come of age like the surface codes that are now shown to work on superconducting platforms and the compact codes that industry heavyweights like Quantinuum are researching the capability to function at physical error rates orders of magnitude below the threshold will enable researchers to construct smaller, yet faster and more efficient, quantum computers. The prospects of quantum computing could very much hinge on how rapidly these record single-qubit fidelities can be replicated in multi-qubit devices, and how successfully the lessons extracted in noise reduction, calibration, and control can be extrapolated to increasingly larger quantum processors.
