Gate-voltage noise limits accuracy of qubit calculations

Researchers Xinning Wang, Bohdan Khromets, Zach Merino and Jonathan Baugh, affiliated with the Institute for Quantum Computing, University of Waterloo, and the Departments of Chemistry and Physics/Astronomy at the University of Waterloo, have identified a critical limitation in silicon spin-qubit technology. Their work demonstrates that gate-voltage noise significantly impacts the accuracy of variational quantum eigensolvers, a specific type of quantum algorithm used for tasks like molecular simulation. The data supporting this study are available upon reasonable request from the authors.

Silicon Spin-Qubits for Variational Quantum Eigensolvers

Silicon spin qubits, a leading candidate for scalable quantum computing, face a critical limitation in their ability to accurately execute variational quantum eigensolvers (VQE), a hybrid quantum-classical algorithm used for molecular simulations. This work links three-dimensional electrostatics to qubit behavior and propagates voltage fluctuations through control pulses, providing a detailed analysis of error sources. The team’s simulations focused on hydrogen’s ground-state energy estimation using VQE as a test case, demonstrating that exchange-based two-qubit gates are approximately ten times more susceptible to gate-voltage noise than single-qubit rotations driven by electron spin resonance.

This disparity stems from the sensitivity of inter-qubit coupling to voltage fluctuations. Further analysis using quantum process tomography and Kraus-operator analysis distinguished between coherent and incoherent error contributions, quantifying the portion of error potentially correctable with a compensating unitary operation.

The study identified specific regimes of gate miscalibration and noise switching time that remain compatible with achieving chemically accurate energy estimates. The researchers modeled dynamically fluctuating charge noise as independent random-telegraph noise on each gate voltage, varying both the amplitude and switching time to simulate different noise spectra. They suggest that statistical post-processing, leveraging the full distribution of noisy energy estimates, could further enhance accuracy. As the paper reports, “We identify regimes of miscalibration strength and noise switching time compatible with chemically accurate energy estimates,” indicating a pathway toward mitigating these errors through algorithmic refinement.

The team developed a co-simulation framework linking three-dimensional electrostatics to qubit behavior, allowing them to propagate voltage fluctuations through realistic control pulses. This sensitivity highlights a specific characteristic within silicon spin-qubit architectures, where inter-qubit coupling is particularly sensitive to external electrical disturbances.

Hardware-Algorithm Co-Simulation Framework Development

Xinning Wang, affiliated with the Institute for Quantum Computing and the Department of Chemistry at the University of Waterloo, led the development of a co-simulation framework designed to model the relationship between hardware imperfections and algorithmic performance in silicon spin qubits. Published on August 12, 2026, the team’s work focuses on the variational quantum eigensolver (VQE) algorithm, utilizing hydrogen’s ground-state energy estimation as a test case to quantify the effects of gate-voltage noise.

The analysis revealed a significant disparity in noise sensitivity between different qubit operations; exchange-based two-qubit gates proved roughly an order of magnitude more susceptible to voltage fluctuations than electron spin resonance-driven single-qubit rotations. This disparity highlights a specific sensitivity within silicon quantum-dot processors, proactively identifying a potential limitation before widespread deployment, and suggesting that addressing noise in two-qubit gate operations will be critical for achieving reliable quantum computation.

Exchange-Based Gates Show Higher Noise Sensitivity

This disparity in noise susceptibility presents a critical challenge for variational quantum eigensolver (VQE) algorithms, a hybrid quantum-classical approach used for molecular simulations and materials science. The results revealed that exchange interactions, fundamental to entangling qubits, are approximately ten times more sensitive to these voltage fluctuations than the single-qubit control mechanisms. This heightened sensitivity wasn’t discovered through abstract modeling; the team constructed a hardware-algorithm co-simulation framework directly linking three-dimensional electrostatics to qubit behavior.

The study specifically investigated the impact on VQE algorithms used for estimating the ground-state energy of the hydrogen molecule (H₂). The work identifies regimes of miscalibration strength and noise switching time compatible with chemically accurate energy estimates, and discusses how statistical post-processing based on the full distribution of noisy energy estimates could further improve accuracy.

Quantum Process Tomography Distinguishes Error Sources

While demonstrating promising coherence times, these qubits exhibit a sensitivity to gate-voltage noise when employed in variational quantum eigensolver (VQE) algorithms. The analysis revealed that the specific structure of the noise, its amplitude and switching time, significantly impacts algorithmic accuracy, particularly within VQE circuits designed to estimate the ground-state energy of molecules like hydrogen. The data supporting this study are available upon reasonable request from the authors.

Random-Telegraph Noise Models Voltage Fluctuations

This approach allowed for the investigation of how both static miscalibration and stochastic fluctuations, modeled as random-telegraph noise with adjustable amplitudes and switching times, impact the accuracy of quantum calculations. Embedding these noise models into the VQE circuit, the researchers identified regimes of miscalibration strength and noise switching time compatible with chemically accurate energy estimates, and discussed how statistical post-processing based on the full distribution of noisy energy estimates could further improve accuracy.

VQE Circuit Testing with H_2 Ground-State Energy

Silicon spin qubits, increasingly investigated for scalable quantum computation, face limitations imposed by gate-voltage noise that directly impacts the accuracy of variational quantum eigensolver (VQE) circuits, according to work published by Xinning Wang, Institute for Quantum Computing, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada; Bohdan Khromets, Department of Chemistry, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada; Zach Merino, Institute for Quantum Computing, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada, and Jonathan Baugh, Department of Physics and Astronomy, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada. The team’s framework links three-dimensional electrostatics to effective g-factors and exchange couplings, allowing for detailed investigation of both static miscalibration and stochastic fluctuations affecting qubit performance. This approach moves beyond abstract noise models to analyze concrete control imperfections within a silicon quantum-dot processor, a crucial step toward realizing practical quantum advantage.

The data supporting this study are available upon reasonable request from the authors. This disparity highlights a specific sensitivity within silicon quantum-dot processors as the technology matures and moves toward more complex quantum algorithms.

Miscalibration and Noise Regimes for Accuracy

They found that the characteristics of the noise, specifically its amplitude and how quickly it switches, significantly influence the reliability of the quantum computation. This detailed understanding of noise behavior is crucial for optimizing qubit control and improving algorithmic performance. The team demonstrated that analyzing the range of possible energy values, rather than relying on a single estimate, can mitigate the impact of noise and yield more reliable results.

Statistical Post-Processing Improves Energy Estimates

This heightened sensitivity stems from the specific characteristics of the noise, particularly its amplitude and the rate at which it fluctuates, influencing the overall fidelity of quantum operations. Understanding these error channels is vital for developing targeted mitigation strategies. Embedding these noise models into the VQE circuit, the authors identify regimes of miscalibration strength and noise switching time compatible with chemically accurate energy estimates, and discuss how statistical post-processing based on the full distribution of noisy energy estimates could further improve accuracy.