Impact and mitigation of Hamiltonian characterization errors in digital-analog quantum computation

Mikel Garcia-de-Andoin, Alatz Álvarez-Ahedo, Adrián Franco-Rubio, Mikel Sanz · 2025

Read Impact and mitigation of Hamiltonian characterization errors in digital-analog quantum computation by Mikel Garcia-de-Andoin, Alatz Álvarez-Ahedo, Adrián Franco-Rubio, Mikel Sanz (2025) 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-2505-03642.txt.

Introduction

Mikel Garcia-de-Andoin, Alatz Álvarez-Ahedo, Adrián Franco-Rubio, and Mikel Sanz study calibration errors in a hybrid style of quantum computing, in a research paper from 2025. The style mixes the device's natural interaction, used as an analog block, with single-qubit digital gates. Its accuracy depends on knowing that natural interaction exactly.

The team bounds how far the implemented interaction can stray from the target one. They also bound how much a measured result can shift because of that straying. Then they propose a fix that resembles echo techniques used to cancel noise. The authors say these bounds and the fix support scaling this approach to larger systems.

Catalog record

CREATORMikel Garcia-de-Andoin, Alatz Álvarez-Ahedo, Adrián Franco-Rubio, Mikel Sanz
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE64 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-03642

Browse the full library · About the Tower of Babel

Full text

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

Impact and mitigation of Hamiltonian characterization errors in digital-analog quantum computation Mikel Garcia-de-Andoin, Alatz Álvarez-Ahedo, Adrián Franco-Rubio, Mikel Sanz Abstract Digital-analog is a universal quantum computing paradigm which employs the natural entangling Hamiltonian of the system and single-qubit gates as resources. Here, we study the stability of these protocols against Hamiltonian characterization errors. For this, we bound the maximum separation between the target and the implemented Hamiltonians. Additionally, we obtain an upper bound for the deviation in the expected value of an observable. We further propose a protocol for mitigating calibration errors which resembles dynamical-decoupling techniques. These results open the possibility of scaling digital-analog to intermediate and large scale systems while having an estimation on the errors committed. I Introduction Quantum computing was first introduced as a way of simulating quantum systems with a controllable device [ 10 , 25 ] . There are two main protocols to control these quantum systems. On the one hand, analog quantum computing (AQC) acts on the system by a continuous variation of some parameters [ 18 ] . Its main advantage is the robustness against noise at the cost of a limited set of control actions over the system. On the other hand, digital quantum computing (DQC) acts on a state with a series of discrete unitary evolutions or gates [ 20 ] . A crucial advantage of DQC is the capacity of implementing a universal set of operations. This allows DQC to implement error correcting codes enabling it to perform simulations without errors [ 2 , 33 , 14 ] . However, the algorithms implemented in DQC are usually less resistant to noise compared to AQC [ 41 , 5 , 49 , 24 ] . Even though we are closer to having access to hardware and techniques allowing this [ 11 , 12 ] , we are still in the noisy intermediate scale quantum devices (NISQ) era [ 42 ] . In this situation, there is a focus on noise mitigation techniques that allow running simulations with less errors [ 15 ] . This would help to fulfill the requirements of the quantum threshold theorem [ 34 , 3 ] and advance to the fault-tolerant era. With this challenge in mind, digital-analog quantum computing (DAQC) [ 40 ] was proposed as a way of achieving the robustness of AQC while maintaining the flexibility of DQC. Instead of trying to mitigate the crosstalk of DQC systems, DAQC employs the natural interaction Hamiltonian as a resource to generate entanglement. Universality of DAQC is achieved by alternating the application of single qubit gates (digital blocks) with periods in which the system freely evolves under the system Hamiltonian for a given time (analog blocks) [ 21 ] . It has been extensively shown that this new paradigm allows the implementation of quantum algorithms [ 38 , 37 , 36 , 44 ] , allowing for efficient circuit designs. The results from [ 29 ] showed a better performance of DAQC against noise compared to similar DQC circuits. Additionally, an error mitigation technique based on zero-noise extrapolation tailored for DAQC was proposed. A great part of the quantum computing community has its focus on developing new tools to reduce the errors in the systems, in order to implement any operation. This is currently done by the engineering of quantum systems and their controls [ 6 , 16 , 32 ] and with the development of active mitigation techniques [ 15 ] such as dynamical-decoupling techniques [ 45 , 23 ] , zero noise extrapolation [ 30 ] , or probabilistic error cancellation [ 48 , 50 ] . The question of reducing the errors up to a threshold to make the quantum devices useful has led to the conclusion that errors must decrease exponentially with the size of the system to perform fault tolerant computations [ 34 , 3 , 47 ] . However, another question can be asked about errors. Is it possible to have a quantum system with such properties that allow us to increase its size without jeopardizing its performance? This question can be answered by looking at the stability of quantum simulation tasks, which can be studied from the error of measuring a certain observable in the thermodynamic limit with constant errors with respect to the system size [ 4 , 51 , 49 , 31 ] . In this article, we study the errors committed in DAQC circuits for quantum simulation tasks under calibration errors in the coupling constants. We first study the difference between the target Hamiltonian and the Hamiltonian simulated with the DAQC circuit. Then, we obtain a bound for the error in the expectation value of an observable. From the insights obtained from these calculations, we develop a new DAQC circuit synthesis method aimed at reducing the effect of the calibration errors. For this, we trade off the total time of the circuit for the stability of the circuit. The rest of the article is organized as it follows. Sec. II introduces the basis for DAQC. In Sec. III we study in detail the stability of DAQC, and Sec. IV follows with the introduction of the error mitigation technique to enhance the stability. In Sec. V we conclude by discussing the applicability of our results to particular experimental setups and we summarize our results in Sec. VI . II Brief introduction to DAQC In general, the goal of a DAQC paradigm is to construct the unitary generated by the evolution under a given problem Hamiltonian H_{\text{P}} for a time T , U=e^{-iTH_{\text{P}}} . The problem Hamiltonian may contain arbitrary single-body terms and two-body interactions. As single-body terms can be implemented through single-qubit gates, our main focus is on the two-body terms {H_{\text{P}}=\sum_{\mu,\nu\in\{x,y,z\}}\sum_{i<j}^{N}{h_{\text{P}}}_{ij}^{\mu\nu}\sigma_{i}^{\mu}\sigma_{j}^{\nu}} . However, a given platform might provide us with a different source Hamiltonian {H_{\text{S}}=\sum_{\mu,\nu\in\{x,y,z\}}\sum_{i<j}^{N}{h_{\text{S}}}_{ij}^{\mu\nu}\sigma_{i}^{\mu}\sigma_{j}^{\nu}} . The task consists in the implementation of U by the sequential application of digital-blocks composed of single-qubit gates (SQG) and analog-blocks of free evolution under the source Hamiltonian H_{\text{S}} for a given prescription of times. For the sake of simplicity, if we assign each Hamiltonian a graph where each qubit is a vertex and the edges are the nonzero couplings of the Hamiltonian labeled by the interaction axes, we will assume that the connectivity graph of the source Hamiltonian must cover the graph of the problem Hamiltonian. Indeed, let \mathcal{S} be the graph associated with H_{\text{S}} and \mathcal{P} to H_{\text{P}} . Then, H_{\text{P}} is directly simulatable if \mathcal{P}\subseteq\mathcal{S} . Otherwise, the graph can be decomposed into several sub-graphs which can be simulated in sequence of alternate SWAP blocks [ 27 ] . To obtain a correct circuit, we make use of the following correspondence, \sigma_{\nu}e^{-iH}\sigma_{\nu}^{\dagger}=e^{-i\sigma_{\nu}H\sigma_{\nu}^{\dagger}} for \sigma_{\nu} a Pauli matrix. If we use the Pauli matrices plus the identity matrix to expand the Hamiltonian and as the applied SQGs, we find that this sandwiching structure allows us to change the signs of some of the terms of H_{\text{S}} , generating what we call an effective Hamiltonian as depicted in Fig. 1 . For simplicity in this paper, we will work in the stepwise-DAQC regime, in which H_{\text{S}} is turned off during the digital blocks. FIG. 1: Representation of a generic stepwise-DAQC

Read the complete text