Refining the process of accurately estimating noisy operations is crucial for building reliable quantum computers, and researchers are now focusing on a technique called gate set tomography (GST). A new study reveals that incorporating “motional degrees of freedom”, the way ions physically move, is the main source of context dependence impacting trapped-ion quantum devices, challenging the long-held assumption of fixed noise levels. Rather than hindering the GST process, the team reports that acknowledging this context dependence and building it into the gate-set parametrization reduces sampling cost. This approach, published in npj Quantum Information, identifies a promising avenue for improving quantum hardware characterization, particularly where microscopic modeling can be applied, and received funding from the US Army Research Office and the European Union’s Horizon Europe research program, including grants such as MILLENION-SGA1 and PID2021-127726NB-I00.
Gate Set Tomography Addresses Non-Markovian Noise in Trapped-Ion Systems
Refining the estimation of noisy quantum gates is increasingly focused on accurately modeling the physical systems themselves, rather than solely improving algorithms. Researchers are now addressing limitations within gate set tomography (GST), a key protocol for characterizing quantum computer performance, by accounting for previously overlooked sources of error in trapped-ion devices. The conventional approach to GST assumes noise operates under “fixed completely-positive and trace-preserving (CPTP) maps,” independent of the sequence of operations applied, a simplification that doesn’t always hold true in real-world experiments. The new study highlights that “motional degrees of freedom” are the primary driver of this discrepancy, introducing context dependence where the order of gates impacts accuracy. This means the noise isn’t static; it evolves based on the history of operations, challenging the Markovian assumption central to standard GST.
The team focused on phonon-mediated two-qubit gates in trapped-ion systems, utilizing detailed microscopic modeling to understand how these interactions introduce context-dependent errors. Understanding these nuances is critical for building reliable quantum hardware capable of complex computations, and is not merely an academic exercise. Incorporating this into the GST parametrization reduces sampling cost, as acknowledging previously ignored factors allows for a more efficient characterization of the system. The researchers report that context dependence can be incorporated into the gate-set parametrization, suggesting a benefit to modeling previously dismissed noise sources. The work, funded by the US Army Research Office through Grant No. W911NF-, and also supported by several European Union grants including MILLENION-SGA1 EU Project and PID2021-127726NB-I00, points toward a context-aware GST as a promising avenue for other quantum platforms where detailed microscopic modeling is feasible, potentially unlocking further improvements in quantum error mitigation and control.
Context-Aware GST Reduces Sampling Cost via Microscopic Modeling
Gate set tomography, or GST, is currently the leading protocol for accurately estimating the performance of noisy quantum gates, state preparations, and measurements, a crucial step toward building practical quantum computers; however, traditional GST methods operate under the assumption of “fixed completely-positive and trace-preserving (CPTP) maps,” a premise increasingly challenged by experimental results. Researchers are now refining GST to account for the reality that noise isn’t static, but evolves based on the sequence of operations applied to a quantum system, a phenomenon known as context dependence. The team addressed this issue through detailed microscopic modeling of high-fidelity light-shift gates, tailoring GST to specifically capture the influence of these motional degrees of freedom. This approach expands GST’s parametrization to acknowledge that noise characteristics can shift depending on the history of gate applications, rather than discarding the core principles of GST. Incorporating context dependence can reduce sampling cost.
The project received funding from the US Army Research Office through Grant No. W911NF-, alongside support from PID2021-127726NB-I00 (MCIU/AEI/FEDER, UE), from the Grant IFT Centro de Excelencia Severo Ochoa CEX2020- -S, funded by MCIN/AEI/ / , from the CSIC Research Platform on Quantum Technologies PTI-001, and from the European Union’s Horizon Europe research and innovation program under grant agreement No (“MILLENION-SGA1” EU Project).
