Universal pulses for superconducting qudit ladder gates

Boxi Li, F. A. Cárdenas-López, Adrian Lupascu, Felix Motzoi · 2024

Read Universal pulses for superconducting qudit ladder gates by Boxi Li, F. A. Cárdenas-López, Adrian Lupascu, Felix Motzoi (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-18339.txt.

Introduction

Li, Cárdenas-López, Lupascu, and Motzoi designed control pulses for qudits in this paper from twenty twenty-four. A qudit is a quantum unit with more than two levels, packing extra information into each site. Higher levels lose their quantum state faster, so gates must be quick and clean. Methods that work for two-level qubits do not extend easily.

The authors give a recipe for pulses that rotate any two neighboring levels. The analysis squeezes errors into a four-level window, counting two leakage levels as well. Their pulses suppress several error types at once and beat prior schemes. Tests on transmon circuits showed strong gains toward very low error rates. The approach should transfer to other ladder-like quantum systems.

Catalog record

CREATORBoxi Li, F. A. Cárdenas-López, Adrian Lupascu, Felix Motzoi
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE79 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-18339

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Universal pulses for superconducting qudit ladder gates Boxi Li, F. A. Cárdenas-López, Adrian Lupascu, Felix Motzoi Abstract Qudits, generalizations of qubits to multi-level quantum systems, offer enhanced computational efficiency by encoding more information per lattice cell, avoiding costly swap operations and providing even exponential speedup in some cases. Utilizing the d d -level manifold, however, requires high-speed gate operations because of the stronger decoherence at higher levels. While analytical control methods have proven effective for qubits in achieving fast gates with minimal control errors, their extension to qudits is nontrivial due to the increased complexity of the energy level structure arising from additional ancillary states. In this work, we present a universal pulse construction for generating rapid, high-fidelity unitary rotations between adjacent qudit levels, thereby providing a prescription for any gate in S ​ U ​ ( d ) SU(d) . Control errors in these operations are effectively analyzed within a four-level subspace, including two leakage levels with approximately opposite detuning. By identifying the optimal degrees of freedom, we derive concise analytical pulse schemes that suppress multiple control errors and outperform existing methods. Remarkably, our approach achieves consistent coherent error scaling across all levels, approaching the quantum speed limit independently of parameter variations between levels. Numerical validation on transmon circuits demonstrates significant improvements in gate fidelity for various qudit sizes aiming for 10 − 4 10^{-4} error. This method provides a scalable solution for improving qudit control and can be broadly applied to other quantum systems with ladder structures or operations involving multiple ancillary levels. I Introduction Quantum computation and quantum information processing protocols often rely on qubits, or two-level systems, as the fundamental units of computation due to their simplicity and close analogy to classical computing. However, most quantum systems comprise more than just two levels. These additional quantum levels can also be used as an information register, which is known as a qudit, a generalization of the qubit to a d -level system. Exploring the full Hilbert space of qudits enables more efficient computation by increasing the amount of information stored per quantum unit. Qudit-based quantum computation offers several known advantages over qubit-based approaches. For instance, a d -level system can encode \log_{2}(d) qubits [ 1 ] . This has been exploited for efficient compilation of arbitrary unitaries, requiring an exponentially reduced number of circuit layers [ 2 , 3 , 4 , 5 , 6 ] , to simulate bosonic modes for studying light-matter processes [ 7 , 8 ] and lattice gauge theories [ 9 , 10 , 11 ] , for enhancing the robustness in quantum cryptography [ 12 , 13 ] , and for simplified implementation of quantum error correction protocols [ 14 , 15 , 16 ] . In general, the larger density of registers means that the connectivity of qubit-based architectures is increased, since neighboring qudits can share up to d^{2} level couplings. Qudit processors have been implemented across various physical platforms, including trapped-ions [ 17 , 18 , 19 , 20 ] , Rydberg atoms [ 21 ] , ultracold atomic mixtures [ 22 ] , molecular spins [ 23 , 24 ] , photonic systems [ 25 , 26 , 27 , 28 , 29 ] and superconducting circuits [ 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 ] ; such implementations contribute to significant progress on qudit-based quantum computation. Despite these advancements, maintaining coherent control of all the qudit levels poses complex new challenges. In transmon superconducting circuits, where the quantum system is represented by a non-linear oscillator [ 38 ] , each qudit operation needs to be addressed differently due to the varying surrounding level structure. Compared to a qubit operation, the presence of additional leakage channels significantly, limits the gate performance, as shown in Fig. 1 a-c. For instance, it has been reported that the gate time of a single-qutrit gate is around 30 ns [ 35 , 33 , 31 ] , which is three times longer than that required for single-qubit gates with state-of-the-art quantum control techniques [ 39 ] . Therefore, developing quantum control protocols for qudits is crucial for making qudit computation practical. Of particular relevance is the Derivative Removal by Adiabatic Gate (DRAG) method [ 40 , 41 , 42 , 43 ] , successfully employed in superconducting qubit systems to reduce leakage and phase errors. DRAG’s simplicity and flexibility allow engineering efficient pulses with easy-to-calibrate parameters, making it ubiquitous in the superconducting qubits platform [ 44 , 45 , 46 , 39 , 47 ] . The same advantages remain even with the presence of multiple error sources, whereby multiple DRAG corrections can be combined, offering efficient yet compact solutions [ 42 , 48 ] . In this article, we investigate the energy structure and dominant error mechanisms of a transmon ladder-type qudit, a weakly anharmonic oscillator, under nearest-level driving. We examine the relationship between the number of usable levels in a transmon qudit and its design parameters, which constrain coherence times and gate design. We demonstrate that transmon qudit control can be studied within a four-level subspace involving two nearest-neighbour interactions and effectively reduced to an enhanced error model in an SU(2) subspace. Through a systematic study across various qudit sizes in transmon circuits, we find that porting the widely-used single-derivative DRAG method, as previously suggested in [ 41 ] and experimentally implemented in [ 32 , 49 ] , does not significantly improve fidelity in the qudit case because it is overconstrained by multiple leakage channels. To overcome this limitation, we apply the DRAG framework to engineer level-independent and high-precision analytical control schemes for qudit systems within a driven ladder structure. We adopt a recursive DRAG approach introduced in Ref. [ 48 ] that incorporates higher-order derivatives, providing new degrees of freedom that are used to suppress both single- and multi-photon errors. Remarkably, despite the presence of multiple parameters in the circuit description, we observe a universal behaviour in the error budget and pulse-specific quantum speed limits, i.e., the minimum gate duration required to achieve a target fidelity. In particular, for a given drive protocol, the speed limit is found to be inversely proportional to the system nonlinearity, and it remains independent of other circuit parameters or the specific qudit transition targeted. However, it is highly sensitive to the drive protocol, particularly to whether certain leakage channels are effectively suppressed. We observe significant improvements in gate performance and successfully reduce gate times to mitigate dephasing caused by voltage fluctuations during gate implementation. These results are broadly applicable to any qudit platform with multiple connected ancillary levels. In the following, we start with the transmon model and derive the four-level effective Hamiltonian in Section II . Next, we introduce and explore the recursive DRAG method in detail and perform a systematic study of its performance in Section III . Significant improvement in fidelity is observed across a wide range of parameters, with a universal behaviour across all levels independent of parameter variations between levels. In Section IV , we discuss other potential control errors beyond the two-level transition and provide a summary of our findings in Section V . II Qudit Model for

Read the complete text