Automated tuning of inter-dot tunnel coupling in double quantum dots

Diepen, C. J. van; Eendebak, Pieter; Buijtendorp, Bruno T.; Mukhopadhyay, U.; Fujita, Takafumi; Reichl, Christian; Wegscheider, Werner; Vandersypen, L. M. K.

doi:10.1063/1.5031034

Cited by 69 publications

(55 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…4F). Here, we use a virtual gate voltage V t12 , where V BC is set while compensating its influence on the dot potentials by appropriate corrections to V P1 and V P2 [6,34]. As a result of this full control over the coupling, we are able to operate the qubits at a mostly charge-insensitive point of symmetric detuning while choosing an exchange coupling strength large enough for rapid two qubit controlled rotations.The advantage of this reduced sensitivity to detuning noise is demonstrated in Fig.…”

mentioning

confidence: 99%

Fast two-qubit logic with holes in germanium

Hendrickx

Franke

Sammak

et al. 2020

Nature

266

283

View full text Add to dashboard Cite

The promise of quantum computation with quantum dots has stimulated widespread research. Still, a platform that can combine excellent control with fast and high-fidelity operation is absent. Here, we show single and two-qubit operations based on holes in germanium. A high degree of control over the tunnel coupling and detuning is obtained by exploiting quantum wells with very low disorder and by working in a virtual gate space. Spin-orbit coupling obviates the need for microscopic elements and enables rapid qubit control with Rabi frequencies exceeding 100 MHz and a singlequbit fidelity of 99.3 %. We demonstrate fast two-qubit CX gates executed within 75 ns and minimize decoherence by operating at the charge symmetry point. Planar germanium thus matured within one year from a material that can host quantum dots to a platform enabling two-qubit logic, positioning itself as a unique material to scale up spin qubits for quantum information.Gate-defined quantum dots were recognized early on as a promising platform for quantum information [1] and a plethora of materials stacks has been investigated as host material. Initial research mainly focused on the low disorder semiconductor gallium arsenide [2,3]. Steady progress in the control and understanding of this system culminated in the initial demonstration and optimization of spin qubit operations [4,5] and the realization of rudimentary analog quantum simulations [6]. However, the omnipresent hyperfine interactions in group III-V materials seriously deteriorate the spin coherence, despite attempts to mitigate this by nuclear polarization [7]. Drastic improvements to the coherence times could be achieved by switching to the group IV semiconductor silicon, in particular when defining spin qubits in isotopically purified host crystal with vanishing concentrations of nonzero nuclear spin [8]. This enabled single qubit rotations with fidelities beyond 99.9% [9] and the execution of two-qubit logic gates with fidelities up to 98% [10-13], underlining the potential of spin qubits for quantum computation. Nevertheless, quantum dots in silicon are often formed at unintended locations and control over the tunnel coupling determining the strength of two-qubit interactions is limited. Moreover, the absence of a sizable spin-orbit coupling for electrons requires the inclusion of microscopic components such as onchip striplines or nanomagnets close to each qubit, which seriously complicates the design of large and dense 2D-structures. This, combined with the limited control over the location and coupling of the dots, remains an outstanding challenge for the scalability of these systems and a platform that can overcome these limitations would be highly desirable. * These two authors contributed equally to this work Hole states in semiconductors typically exhibit strong spinorbit coupling, which has enabled the demonstration of fast single qubit rotations [14,15] and additionally, unlike electrons, holes do not suffer from nearby valley states. In silicon, unfavorable band alignment...

show abstract

mentioning

confidence: 99%

Fast two-qubit logic with holes in germanium

Hendrickx

Franke

Sammak

et al. 2020

Nature

266

283

View full text Add to dashboard Cite

show abstract

“…Progress in quantum dots array sizes has been steady. Given that double and triple dots are being routinely used in experiments [17][18][19] and moderate linear array sizes (∼ 10 dots) [7], as well as two-dimensional arrays [6], are on the horizon, an auto-tuning procedure for these devices is a significant step for employing them in both the laboratories and in applications.…”

Section: Resultsmentioning

confidence: 99%

QFlow lite dataset: A machine-learning approach to the charge states in quantum dot experiments

Zwolak

Kalantre

et al. 2018

PLoS ONE

View full text Add to dashboard Cite

BackgroundOver the past decade, machine learning techniques have revolutionized how research and science are done, from designing new materials and predicting their properties to data mining and analysis to assisting drug discovery to advancing cybersecurity. Recently, we added to this list by showing how a machine learning algorithm (a so-called learner) combined with an optimization routine can assist experimental efforts in the realm of tuning semiconductor quantum dot (QD) devices. Among other applications, semiconductor quantum dots are a candidate system for building quantum computers. In order to employ QDs, one needs to tune the devices into a desirable configuration suitable for quantum computing. While current experiments adjust the control parameters heuristically, such an approach does not scale with the increasing size of the quantum dot arrays required for even near-term quantum computing demonstrations. Establishing a reliable protocol for tuning QD devices that does not rely on the gross-scale heuristics developed by experimentalists is thus of great importance.Materials and methodsTo implement the machine learning-based approach, we constructed a dataset of simulated QD device characteristics, such as the conductance and the charge sensor response versus the applied electrostatic gate voltages. The gate voltages are the experimental ‘knobs’ for tuning the device into useful regimes. Here, we describe the methodology for generating the dataset, as well as its validation in training convolutional neural networks.Results and discussionFrom 200 training sets sampled randomly from the full dataset, we show that the learner’s accuracy in recognizing the state of a device is ≈ 96.5% when using either current-based or charge-sensor-based training. The spread in accuracy over our 200 training sets is 0.5% and 1.8% for current- and charge-sensor-based data, respectively. In addition, we also introduce a tool that enables other researchers to use this approach for further research: QFlow lite—a Python-based mini-software suite that uses the dataset to train neural networks to recognize the state of a device and differentiate between states in experimental data. This work gives the definitive reference for the new dataset that will help enable researchers to use it in their experiments or to develop new machine learning approaches and concepts.

show abstract

“…A ML algorithm treats any given task as a mathematical problem and does not utilize knowledge of the specific physics underlying the data. Benefiting from this 'model-free' nature, the ML algorithms find broad application in various sub-fields of physics, such as the identification of phase transitions in condensed matter studies [31,32], the classification of multi-qubit states of trapped-ion experiments [33], the auto-tuning of gate voltages in quantum dots system [34,35], and the future state predictions of spatio-temporal chaotic systems [36][37][38]. Although successfully applied in various studies, one crucial drawback of the ML techniques is the trade-off between a successfully trained program and the amount of data required during its training phase.…”

Section: Introductionmentioning

confidence: 99%

Classification and Prediction of Wave Chaotic Systems with Machine Learning Techniques

Xiao

Hong

et al. 2019

Acta Phys. Pol. A

View full text Add to dashboard Cite

The wave properties of complex scattering systems that are large compared to the wavelength, and show chaos in the classical limit, are extremely sensitive to system details. A solution to the wave equation for a specific configuration can change substantially under small perturbations. Due to this extreme sensitivity, it is difficult to discern basic information about a complex system simply from scattering data as a function of energy or frequency, at least by eye. In this work, we employ supervised machine learning algorithms to reveal and classify hidden information about the complex scattering system presented in the data. As an example we are able to distinguish the total number of connected cavities in a linear chain of weakly coupled lossy enclosures from measured reflection data. A predictive machine learning algorithm for the future states of a perturbed complex scattering system is also trained with a recurrent neural network. Given a finite training data series, the reflection/transmission properties can be forecast by the proposed algorithm. arXiv:1908.04716v1 [cond-mat.dis-nn]

show abstract

Automated tuning of inter-dot tunnel coupling in double quantum dots

Cited by 69 publications

References 17 publications

Fast two-qubit logic with holes in germanium

Fast two-qubit logic with holes in germanium

QFlow lite dataset: A machine-learning approach to the charge states in quantum dot experiments

Classification and Prediction of Wave Chaotic Systems with Machine Learning Techniques

Contact Info

Product

Resources

About