Scalable, high-fidelity all-electronic control of trapped-ion qubits

TL;DR

The paper tackles the scalability and noise challenges in trapped-ion quantum computing by proposing an all-electronic architecture that uses a shared AC magnetic drive together with local DC tuning electrodes for site-selective, laser-free gate control. It validates the approach experimentally in a seven-zone trap, achieving record fidelities such as single-qubit gate fidelity around 0.9999916 with small uncertainties and two-qubit entanglement fidelity around 0.9997, with sustained performance over extended operation. The authors describe the effective control mechanisms and demonstrate precise site-selectivity and low crosstalk, along with tunable-angle two-qubit gates and long-term stability, indicating a practical path to robust large-scale TIQCs. This work suggests a scalable route to thousands of qubits on standard microfabrication platforms, reducing laser infrastructure and enabling efficient multiplexed electronic control for future quantum processors.

Abstract

The central challenge of quantum computing is implementing high-fidelity quantum gates at scale. However, many existing approaches to qubit control suffer from a scale-performance trade-off, impeding progress towards the creation of useful devices. Here, we present a vision for an electronically controlled trapped-ion quantum computer that alleviates this bottleneck. Our architecture utilizes shared current-carrying traces and local tuning electrodes in a microfabricated chip to perform quantum gates with low noise and crosstalk regardless of device size. To verify our approach, we experimentally demonstrate low-noise site-selective single- and two-qubit gates in a seven-zone ion trap that can control up to 10 qubits. We implement electronic single-qubit gates with 99.99916(7)% fidelity, and demonstrate consistent performance with low crosstalk across the device. We also electronically generate two-qubit maximally entangled states with 99.97(1)% fidelity and long-term stable performance over continuous system operation. These state-of-the-art results validate the path to directly scaling these techniques to large-scale quantum computers based on electronically controlled trapped-ion qubits.

Figures (3)

• Figure 1: Illustration of the all-electronic approach to the coherent control of trapped-ion qubits. The TIQC is a microfabricated chip (center) that contains a repeating 2D grid of unit cells (left), each composed of trapped-ion qubits stored near local tuning electrodes. A shared drive is routed to deliver AC magnetic fields to all unit cells at once. The strength of qubit coupling with the shared drive is controlled by applying DC electric fields to the tuning electrodes. Right: a linear seven-zone test device used in experiments in section \ref{['sec:experiments']}. The tuning electrodes and a shared drive are placed on the same layer. The shared current-carrying trace is routed along $\hat{x}$, and the ions are trapped directly above at the height of $z \approx 40 \mathrm{\mu m}$. Laser light is delivered to zone 4 for ion loading, state preparation, and measurement, but is not used for coherent control. A static magnetic field $\mathrm{B_0} \approx 8.5$ mT is oriented along $\hat{y}$.
• Figure 2: Summary of single-qubit control experiments. A) Single-qubit control is achieved by using DC electrodes to adjust qubit position along $\hat{y}$. Directly above the shared trace ($y \approx 0$) the AC magnetic field is oriented in the $y$-direction. Since the static magnetic field is oriented along $y$ as well, we find that the single-qubit Rabi frequency $\Omega_1 \propto B_z \approx 0$, i.e. the qubit is "hidden" (appendix \ref{['secA1']}). At the same time, $|\partial \Omega_1 / \partial y| \propto |\partial B_z / \partial y| > 0$, thus translating the qubit along $y$ causes interactions to be turned on. B) (Top left) Experimental measurements of the single-qubit Rabi frequency $\Omega_1$ vs $y$. (Right) Rabi oscillations in the "active" (top) and "hidden" (bottom) positions. (Bottom left) Spin-flip probability $P(\downarrow)$ vs $y$ for an SK1 sequence calibrated to drive a $\pi$-pulse in the "active" position. c) Results of single-qubit RB throughout the device. Each experiment uses one qubit to record the spin-flip probability $P(\downarrow)$ in a target zone in one of three configurations: all zones "active" (top), target zone "active" and other zones hidden (middle), and all zones "hidden" (bottom). The exponential fits (dashed lines) are used to estimate the error per Clifford (epc) in each configuration.
• Figure 3: Summary of two-qubit control experiments. A) Two-qubit Molmer-Sorensen interactions are implemented by driving currents through the shared trace. A state-dependent force on the in-plane radial mode is generated using the magnetic field gradient $\partial B_z / \partial y$, and the fact that $B_z \approx 0$ is used to minimize off-resonant qubit frequency shifts. Quadrupole potentials generated by the DC electrodes are used to tune the motional mode frequencies and orientations, adjusting the effective interaction strength. B) The entangled state fidelity in zone 4 following full system calibration. We record a fidelity of $0.9997(1)$ over the first 12 hours, and a fidelity of $0.9995(1)$ over the full 60 hours of data acquisition. The $68\%$ confidence intervals (CI) are calculated by bootstrapping with 10,000 resamples. After 12 hours, the CI deviates visibly from the $1/\sqrt{t}$ scaling (inset), indicating drifts in the underlying process. C) Qubit parity oscillations in 5 zones (not corrected for state preparation and measurement errors), demonstrating the ability to prepare maximally entangled states $\theta_2 \approx \pi/2$ throughout the device. D) Experimental measurements of the entangling angle $\theta_2$ when the mode frequency is electrically adjusted to $n \times \delta$, where $\delta$ corresponds to a detuning of a maximally entangling operation. Insert shows parity oscillations used to infer the value of $\theta_2$.