Pith. sign in

REVIEW 4 major objections 5 minor 107 references

NeuroQD: A Learning-Based Simulation Framework For Quantum Dot Devices

T0 review · 4 major / 5 minor · reviewed 2026-08-05 · deepseek-v4-flash

Pith's one-line read NeuroQD claims that the slow electrostatic solver in quantum-dot device simulation can be replaced by a compact CNN, because the gate-voltage-to-2DEG mapping is a heterostructure-defined blur that transfers from a 2-dot training device to 9

desk verdict NeuroQD is a genuinely useful surrogate for COMSOL electrostatics in Si/SiGe QD devices, but the headline accuracy number covers only the raw potential, not the post-processed charge states the simulator is meant to deliver. read the letter →

arxiv 2509.02872 v1 pith:2OQFRJZ6 submitted 2025-09-02 cond-mat.mes-hall

classification cond-mat.mes-hall
keywords quantumdotdevicesspinqubitsdevicesimulationconvolutionalneuralnetworkU-Net2DEGpotentialreal-timecontrolsurrogatemodel
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proposes a simulator for quantum-dot spin-qubit devices that is both physically realistic and fast enough to run inside a live experiment-control loop: a compact convolutional network (U-Net) learns the mapping from gate voltages to the electrostatic potential in the qubit layer (the 2DEG), replacing the slow finite-element solver that would normally compute it. The key claim is that this mapping is essentially a distance-dependent blur set by the wafer's layer stack, not by the gate layout, so a network trained once on a tiny two-dot device keeps >96% agreement with the physics-based COMSOL baseline when applied to one-dimensional arrays of up to 99 dots, while running 1000x faster. Integrated into a real control stack, the simulator reproduces the turn-on curve, Coulomb peaks, and charge stability diagrams observed on devices operated at 9 mK, with millisecond latency. The paper's practical goal is to let tuning software and control hardware be developed and tested in the loop without risking damage to fragile devices.

What carries the argument

The load-bearing object is the learned blur: the paper models the propagation of the gate-layer potential through the wafer to the 2DEG layer as a fully-convolutional blurring operator whose kernel is set by the heterostructure (Si cap, SiGe spacer, Si quantum well). This operator is realized by a U-Net with skip connections, trained on 18k COMSOL-generated 2DEG potentials for a 2-dot device (augmented to ~148k samples). The generalization step is carried by the translation invariance and input-size agnosticism of the fully convolutional architecture: since the blur kernel does not depend on where a gate sits or how many there are, a model that has learned the local blur on a small device tr

What would settle it

Compare the trained 2-dot U-Net to COMSOL ground truth for a same-heterostructure device in which the gate pitch is doubled; if 2DEG-potential accuracy falls below the >96% bound, the blurring operator is not layout-independent and the transfer claim fails. A complementary check is to vary the 2DEG depth (wafer) by a few nanometers: the blur kernel should change, so accuracy should drop; if it does not, the model is learning something other than the heterostructure-defined blur.

Watch

Extended reading notes

Core claim

The central discovery is Observation 2: the transformation from the gate-layer potential (the pattern of voltages on the top surface) to the 2DEG potential 59 nm below is a high-order blurring operation, a fully convolutional transformation that is determined by the heterostructure alone and is independent of the gate layout. Because convolutional networks are translation invariant and input-size agnostic, a U-Net trained on a single 2-dot device can infer the 2DEG potential for larger devices built on the same wafer stack. Armed with this surrogate, the paper builds a complete simulator: gate voltages are painted onto gate regions to form the input, the U-Net predicts the 2DEG potential, an

Load-bearing premise

The claim that a two-dot-trained network generalizes to 99 dots rests on the assumption that the layer stack blurs nearby voltages the same way everywhere, so no new global or long-range effect appears when the device grows.

Editorial extensions

If this is right

  • Control software and autotuning algorithms can be developed against a realistic device model that answers in milliseconds, the same timescale as the actual set-wait-read loop, so bugs that would destroy a real chip can be caught in simulation.
  • The 100% convergence of the surrogate removes the failure mode of the finite-element solver (which failed on roughly 8–10% of test voltage configurations), making exhaustive sweeps of a device's operating range feasible.
  • Because runtime scales linearly with the number of dots (up to 99qd in this paper), large 1xN arrays that are currently impractical for COMSOL become simulable for architecture exploration.
  • The simulator reproduces measurable device physics (turn-on threshold, Coulomb-peak spacing, charge-stability line slopes), so it can serve as a testbed for tuning procedures before hardware is available.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The blurring-operator claim, if true for 1xN arrays, plausibly extends to 2D dot arrays on the same heterostructure, but the paper does not demonstrate this; 2D layouts introduce corners, more varied cross-coupling, and boundary effects that could violate the strict translation invariance. A natural extension is to train on a 2x2 cell and test on larger 2xN arrays.
  • The >96% accuracy is measured on the 2DEG potential itself; accuracy of derived quantities (charge transition positions, sensor currents) could in principle degrade differently. The real-device validation covers a 2-dot region only, so a follow-up would calibrate how potential-space error bounds translate into tuning-relevant observables on larger devices.
  • A sharper test of the heterostructure-determinism claim would be to train on one wafer stack and apply to another with the same gate layout but a slightly different 2DEG depth: if the blur kernel is truly stack-defined, accuracy should drop and retraining on a single small device should recover it, confirming the parameterization is minimal.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes NeuroQD, a machine-learning surrogate for COMSOL electrostatics simulation of Si/SiGe quantum dot devices. A U-Net is trained on 2-dot COMSOL data to map a gate-layer potential (constructed from gate voltages and layout) to the 2DEG potential, exploiting the observation that this mapping is a fully-convolutional 'blurring' operation. The authors report >96% sMAPE agreement with COMSOL on devices with 2–12 dots, >1000x speedup, apparent generalization to 99-dot devices, and integration into a real-time experimental control stack. The integrated simulator produces turn-on curves, slit-vs-sensor sweeps, Coulomb peaks, and charge stability diagrams that qualitatively match two real devices operated at 9 mK, with comparable charging energies and cross-capacitance slopes. The post-processing chain includes a 2DEG electron-density model, a charge-state integration with an added neighboring-electron energy penalty, and a simulated sensor current.

Significance. If the central claims were fully supported, NeuroQD would be a practically useful tool for architecture research and tuning-software development for spin-qubit arrays, where real-time device-level simulation is currently missing. The paper has genuine strengths: the training is performed on the smallest 2-dot device with a transparent sampling strategy (LHS plus graph-based cross-coupling sampling), the generalization to larger 1xN arrays of the same heterostructure is empirically demonstrated, the speedup is real, and the comparison to two in-house fabricated devices is a step beyond purely synthetic benchmarks. However, the validation currently does not establish the fidelity of the simulator's actual outputs (charge states, sensor currents) because the quantitative accuracy metric is computed on the raw 2DEG potential, while the post-processed observables are compared only qualitatively and depend on hand-set parameters. The significance of the work is therefore conditional on additional end-to-end validation.

major comments (4)
  1. [§7.1.4, Table 4; §6.3.1] This is a specific technical concern that undermines the central claim of '>96% agreement' as applied to the simulator's actual outputs.
  2. [§6.3.1, §7.2] This is a load-bearing issue because the only independent physical check of the simulator is the real-device comparison, and the parameters entering that comparison are hand-set.
  3. [§5.3, §7.1.5] This concerns the central scalability claim and needs either additional evidence or a scope restriction.
  4. [Table 3, §7.1.3] This is a bias concern in the central accuracy comparison.
minor comments (5)
  1. [§6.1.2] Typo: 'electrary charge' should be 'elementary charge.'
  2. [§7.2.4] Typo: 'Figrue 14' should be 'Figure 14.'
  3. [§3] Typo: 'heterosteucture' should be 'heterostructure.'
  4. [Figure 16] The COMSOL runtime is an estimate (average per-configuration time multiplied by the number of configurations), not a measured sweep runtime. Please state this clearly in the caption and, if possible, include a measured runtime for a few representative COMSOL configuration runs.
  5. [§7.1.4] The definition of 'Accuracy = 1 − sMAPE' is nonstandard because sMAPE is already an average over grid points; clarify whether Accuracy is computed per-sample and then averaged, or globally over all points, and how the reported per-category and overall numbers relate to Table 4.

Circularity Check

1 steps flagged · score 3.0 of 10

One disclosed post-processing parameter is fitted to reproduce a feature that is later cited as experimental agreement; the central U-Net generalization claim itself is not circular.

  1. fitted input called prediction [Section 6.3.1 (Charge State Model) and Section 7.2.5 (Charge Stability Diagram)]
    "To simulate the triple-point feature observed in charge stability diagrams, we applied a constant energy penalty for each electron already present in neighboring dots when adding an additional electron to the dot."

    The neighbor-electron energy penalty is introduced specifically to make the simulated charge stability diagram exhibit the triple-point feature seen in experiments. Section 7.2.5 then cites the resulting 'defined charge state regions ... matching the shapes observed in Devices 1 and 2' as evidence that the simulator reproduces real-device behavior. For the triple-point shape, the agreement is therefore by construction: the feature was put into the post-processing model, not independently predicted by the U-Net potential or derived from COMSOL. Other CSD metrics (charging energy spacing, cross-capacitance slopes) are not fixed by this penalty and retain independent content, so the circularity is partial.

full rationale

The core contribution—training a U-Net on 2-dot COMSOL data and testing on larger COMSOL devices—is a standard surrogate-model evaluation, not circular: the test configurations and device sizes are outside the training set, and the model could have failed. The >96% agreement is measured against the same COMSOL family that produced the training data, but that is a benchmark choice, not a reduction of the prediction to its inputs. The generalization to 99 dots is empirically tested, and the claim that the gate-layer-to-2DEG transformation is layout-independent is a testable physical observation, not a definitional tautology. The real-device comparisons provide external evidence, though qualitative. The one genuine circular step is the constant energy penalty in Section 6.3.1, which is added specifically to reproduce the triple-point feature and later cited as matching experiment; this makes part of the CSD agreement by construction. The V_on parameter may also be chosen to align the turn-on jump, but the paper does not explicitly state that it is fitted, so it is not counted here. No load-bearing self-citation chain or imported uniqueness theorem was found. Overall, the central claim is independent, with one disclosed post-processing fit producing a partial circularity.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The central claim rests on three types of unpaid inputs: the layout-independence and bounded-voltage assumptions, the simplified ideal 2DEG physics model, and the COMSOL ground truth. No new physical entities are introduced. The main free parameters are the turn-on potential, the hand-set energy penalty for neighboring electrons, and an unspecified sensor-current calibration, all of which influence how well simulated device metrics match real measurements.

free parameters (3)
  • V_on (turn-on potential) = not stated in paper
    Used in the ideal 2DEG density model n2D(V) in Section 6.1.2 and needed to match the turn-on threshold observed in real devices; set as a constant but not derived from first principles.
  • Neighboring-electron energy penalty = not stated in paper
    Introduced in Section 6.3.1 as a constant energy penalty for each electron already present in neighboring dots when adding an electron; it is hand-set to reproduce triple-point features in charge stability diagrams.
  • Sensor current conversion scale = not stated in paper
    The post-processing section says it computes sensor dot current but does not give the model or calibration constants used to convert electron densities into the simulated demodulated signal plotted against real devices.
assumptions (6)
  • domain assumption The gate-layer to 2DEG potential transformation is determined by the heterostructure alone and is independent of gate layout.
    Observation 2 in Section 5.3; load-bearing for the claim that a model trained on 2qd generalizes to 99qd. If this assumption fails, the main generalization result is unsupported.
  • domain assumption Gate voltages are confined to the safe range 0 to 1 V for the in-house Si/SiGe devices.
    Observation 1 in Section 5.1; justifies restricting the sampling and inference domain. This is a device-specific safety constraint, not a general physical law.
  • domain assumption The ideal 2DEG model at T = 0 K with piecewise linear density n2D(V) = (m* e (V - V_on)/(pi hbar^2)) for V >= V_on, else 0, adequately represents device electrostatics.
    Section 6.1.2; this simplified model is encoded in COMSOL and used in post-processing. It ignores finite temperature, subband structure beyond the first subband, and quantum confinement effects beyond a simple density cutoff.
  • domain assumption COMSOL electrostatics solutions are a suitable ground truth for the 2DEG potential in real devices.
    Section 3; the surrogate is trained and evaluated against COMSOL. The paper cites prior works for COMSOL applicability, but this is still an external modeling assumption inherited from those references.
  • domain assumption The gate-layer to 2DEG transformation is invariant under rotation and reflection of the gate layout.
    Appendix A; used to justify 8x data augmentation. If the heterostructure or gate geometry breaks this symmetry, the augmented samples would be invalid.
  • standard math U-Net is translation invariant and input-size agnostic.
    Section 5.3, citing Long et al. [56]; used to argue the trained CNN can be applied to larger input grids. This is a standard property of fully convolutional networks.

how reviews work

0 comments
Cite this review

Pith. "Pith review of NeuroQD: A Learning-Based Simulation Framework For Quantum Dot Devices." pith.science (2026). https://pith.science/paper/2OQFRJZ6

@misc{pith2026250902872,
  author       = {Pith},
  title        = {Pith review of: NeuroQD: A Learning-Based Simulation Framework For Quantum Dot Devices},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2OQFRJZ6}},
  note         = {Machine review of arXiv:2509.02872}
}
read the original abstract

Electron spin qubits in quantum dot devices are promising for scalable quantum computing. However, architectural support is currently hindered by the lack of realistic and performant simulation methods for real devices. Physics-based tools are accurate yet too slow for simulating device behavior in real-time, while qualitative models miss layout and wafer heterostructure. We propose a new simulation approach capable of simulating real devices from the cold-start with real-time performance. Leveraging a key phenomenon observed in physics-based simulation, we train a compact convolutional neural network (CNN) to infer the qubit-layer electrostatic potential from gate voltages. Our GPU-accelerated inference delivers >1000x speedup with >96% agreement to the physics-based simulation. Integrated into the experiment control stack, the simulator returns results with millisecond scale latency, reproduces key tuning features, and yields device behaviors and metrics consistent with measurements on devices operated at 9 mK.

Figures

Figures reproduced from arXiv: 2509.02872 by the authors.

Figure 1
Figure 1. Overview of the target quantum dot device simu￾lation Different from other works [41, 43, 44, 80] on quantum pro￾gram or Hamiltonian simulation, this paper targets device￾level simulation of QD devices and its charge-sensor re￾sponse to applied gate voltages. As illustrated in [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Schematic of Si/SiGe heterostructure. Gates are fabricated on top and the electrons are trapped in the middle layer. 2 Background 2.1 Quantum Dot Devices Quantum dot devices are semiconductor devices that can trap and control individual electrons to encode qubits [10, 32, 39, 57, 87]. The trapping of individual electrons is achieved by providing confinement potential in both the vertical di￾rection and the lateral d… view at source ↗
Figure 4
Figure 4. Overall approach modeling of these devices in COMSOL and compare the sim￾ulation to experiments. While gate architectures that imple￾ment 2×𝑛 and 𝑛×𝑛 dot arrays were demonstrated (more com￾mon on Ge/SiGe and SiMOS heterostructrues) [27, 35, 82, 99], 1 × 𝑛 dot array remains the most mature device architecture as of today [68]. 4 Overall Approach The overall approach taken by NeuroQD can be described as identifying th… view at source ↗
Figures from the paper (13 more)
Figure 5
Figure 5. Figure 5: Gate-layer potential from gate layout In the gate layer (the top surface of the wafer), each gate defines an electrostatic potential in the region directly be￾neath it. As a result, the gate-layer potential is partitioned into disjoint regions, each governed by the vol…
Figure 6
Figure 6. Figure 6: Potential profile at different depths as it traverses down the dielectrics. We model this blurring effect using CNN. 2DEG layer, the original structure has been smoothed into a continuous, soft-edged profile. This transformation resem￾bles a high-order blurring operati…
Figure 7
Figure 7. Figure 7: Device geometry based on gate layout and het￾erostructure the gate-layer potential to the 2DEG potential. Once trained, the model serves as a fast, learned approximation of the COM￾SOL simulation step, which eliminates the computational bottleneck in the baseline. Fina…
Figure 8
Figure 8. Figure 8: Graph for extracting cross-coupled subspaces and serve as a proxy for cross-coupling. This graph ( [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 9
Figure 9. Figure 9: Experiment setup, server, and real-time simulation backend simulator returns the most recent simulated demodulated signal. 7 Evaluation In this section, we evaluate NeuroQD by 1) comparing the accuracy and performance of the trained U-Net model to the COMSOL simulation…
Figure 10
Figure 10. Figure 10: Comparison against the COMSOL baseline 95 100 UNet accuracy(%) 1000 2000 COMSOL Runtime 2 3 4 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75 78 81 84 87 90 93 96 99 Device (qd) 0 10 UNet Runtime Gap: 15-900 s runtime (s) [PITH_FULL_IMAGE:figure…
Figure 11
Figure 11. Figure 11: Accuracy and runtime trends 7.1.4 Running Test Samples on U-Net Model. We bench￾mark the accuracy of the U-Net model using the converged test samples from COMSOL. Accuracy is evaluated using the symmetric mean absolute percentage error (sMAPE), defined as: sMAPE = 1 𝑛…
Figure 12
Figure 12. Figure 12: , these features were observed on both devices and simulated by the simulation. Simulation 0.0010 0.0008 0.0006 0.0002 0.0004 0.0000 Device1 Turn-on Jump Saturation Device2 Turn-on Jump Saturation Turn-on Jump Saturation 0.0000 0.0002 0.0006 0.0008 0.0010 0.0012 0.000…
Figure 15
Figure 15. Figure 15: indicate the number of electrons trapped in the dot under P1 and P2. Two key physical quantities that we can compare in this experiment are the charging energy, as defined by the spacing between the transition lines (yellow arrows in [PITH_FULL_IMAGE:figures/full_fig…
Figure 16
Figure 16. Figure 16: Runtime comparison 11 [PITH_FULL_IMAGE:figures/full_fig_p011_16.png]
Figure 17
Figure 17. Figure 17: (a) Gate-layer potential approximation, (b) Train￾ing data augmentation via rotation and reflection Data Augmentation via Rotation and Reflection We apply geometric transformations to the input/output pairs. The target transformation is invariant under rotation and re…
Figure 18
Figure 18. Figure 18: summarizes the layout of the 35 devices (6 near￾term devices and 29 extended devices) used in the experi￾ments in Section 7.1 [PITH_FULL_IMAGE:figures/full_fig_p015_18.png]
Figure 19
Figure 19. Figure 19: Device 1 and device 2 dose test C U-Net architecture [PITH_FULL_IMAGE:figures/full_fig_p016_19.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

107 extracted references · 74 canonical work pages

  1. [1]

    [n. d.]. COMSOL® Software Version 6.3 Release Highlights. https: //www.comsol.com/release/6.3 [Online; accessed 2025-06-20]

  2. [2]

    Christopher R Anderson, Mark F Gyure, Sam Quinn, Andrew Pan, Richard S Ross, and Andrey A Kiselev. 2022. High-precision real- space simulation of electrostatically confined few-electron states. AIP Advances 12, 6 (2022)

  3. [3]

    Hidehiro Asai, Shota Iizuka, Tsutomu Ikegami, Junichi Hattori, Koichi Fukuda, Hiroshi Oka, Kimihiko Kato, Hiroyuki Ota, and Takahiro Mori. 2021. Development of integrated device simulator for quan- tum bit design: self-consistent calculation for quantum transport and qubit operation. In 2021 5th IEEE Electron Devices Technology & Manufacturing Conference ...

  4. [4]

    Timothy A Baart, Pieter T Eendebak, Christian Reichl, Werner Wegscheider, and Lieven MK Vandersypen. 2016. Computer- automated tuning of semiconductor double quantum dots into the single-electron regime. Applied Physics Letters 108, 21 (2016)

  5. [5]

    Ali Bakhoda, George L Yuan, Wilson WL Fung, Henry Wong, and Tor M Aamodt. 2009. Analyzing CUDA workloads using a detailed GPU simulator. In 2009 IEEE international symposium on performance analysis of systems and software . IEEE, 163–174

  6. [6]

    Félix Beaudoin, Pericles Philippopoulos, Chenyi Zhou, Ioanna Kriek- ouki, Michel Pioro-Ladrière, Hong Guo, and Philippe Galy. 2022. Robust technology computer-aided design of gated quantum dots at cryogenic temperature. Applied physics letters 120, 26 (2022)

  7. [7]

    Nathan Binkert, Bradford Beckmann, Gabriel Black, Steven K Rein- hardt, Ali Saidi, Arkaprava Basu, Joel Hestness, Derek R Hower, Tushar Krishna, Somayeh Sardashti, et al. 2011. The gem5 simulator. ACM SIGARCH computer architecture news 39, 2 (2011), 1–7

  8. [8]

    Stefan Birner, Tobias Zibold, Till Andlauer, Tillmann Kubis, Matthias Sabathil, Alex Trellakis, and Peter Vogl. 2007. Nextnano: general purpose 3-D simulations. IEEE Transactions on Electron Devices 54, 9 (2007), 2137–2142

Show all 107 references
  1. [9]

    F Borjans, X Mi, and JR Petta. 2021. Spin digitizer for high-fidelity readout of a cavity-coupled silicon triple quantum dot. Physical Review Applied 15, 4 (2021), 044052

  2. [10]

    Guido Burkard, Thaddeus D Ladd, Andrew Pan, John M Nichol, and Jason R Petta. 2023. Semiconductor spin qubits. Reviews of Modern Physics 95, 2 (2023), 025003

  3. [11]

    Zhenyu Cai, Michael A Fogarty, Simon Schaal, Sofia Patomäki, Si- mon C Benjamin, and John JL Morton. 2019. A silicon surface code architecture resilient against leakage errors. Quantum 3 (2019), 212

  4. [12]

    Anasua Chatterjee, Paul Stevenson, Silvano De Franceschi, Andrea Morello, Nathalie P de Leon, and Ferdinand Kuemmeth. 2021. Semi- conductor qubits in practice. Nature Reviews Physics 3, 3 (2021), 157–177

  5. [13]

    Shize Che, Seongwoo Oh, Haoyun Qin, Yuhao Liu, Anthony Sigillito, and Gushu Li. 2024. Fast Virtual Gate Extraction For Silicon Quantum Dot Devices. In Proceedings of the 61st ACM/IEEE Design Automation Conference. 1–6

  6. [14]

    Mingkun Chen, Robert Lupoiu, Chenkai Mao, Der-Han Huang, Jiaqi Jiang, Philippe Lalanne, and Jonathan A Fan. 2022. High speed simulation and freeform optimization of nanophotonic devices with physics-augmented deep learning. ACS Photonics 9, 9 (2022), 3110– 3123

  7. [15]

    Jesus D Cifuentes, Philip Y Mai, Frédéric Schlattner, H Ekmel Ercan, MengKe Feng, Christopher C Escott, Andrew S Dzurak, and Andre Saraiva. 2023. Path-integral simulation of exchange interactions in cmos spin qubits. Physical Review B 108, 15 (2023), 155413

  8. [16]

    Jesús D Cifuentes, Tuomo Tanttu, Will Gilbert, Jonathan Y Huang, En- sar Vahapoglu, Ross CC Leon, Santiago Serrano, Dennis Otter, Daniel Dunmore, Philip Y Mai, et al. 2024. Bounds to electron spin qubit variability for scalable CMOS architectures. Nature communications 15, 1 (...

  9. [17]

    Stefanie Czischek, Victor Yon, Marc-Antoine Genest, Marc-Antoine Roux, Sophie Rochette, Julien Camirand Lemyre, Mathieu Moras, Michel Pioro-Ladrière, Dominique Drouin, Yann Beilliard, et al. 2021. Miniaturizing neural networks for charge state autotuning in quan- tum dots. Mac...

  10. [18]

    Jana Darulová, SJ Pauka, Nathan Wiebe, Kok W Chan, GC Gardener, Michael J Manfra, Maja C Cassidy, and Matthias Troyer. 2020. Au- tonomous tuning and charge-state detection of gate-defined quantum dots. Physical Review Applied 13, 5 (2020), 054005

  11. [19]

    Carl De Boor. 1972. On calculating with B-splines. Journal of Ap- proximation theory 6, 1 (1972), 50–62

  12. [20]

    JP Dodson, Nathan Holman, Brandur Thorgrimsson, Samuel F Neyens, ER MacQuarrie, Thomas McJunkin, Ryan H Foote, LF Edge, SN Coppersmith, and MA Eriksson. 2020. Fabrication process and failure analysis for robust quantum dots in silicon. Nanotechnology 31, 50 (2020), 505001

  13. [21]

    Jingyu Duan, Michael A Fogarty, James Williams, Louis Hutin, Maud Vinet, and John JL Morton. 2020. Remote capacitive sensing in two-dimensional quantum-dot arrays. Nano Letters 20, 10 (2020), 7123–7128

  14. [22]

    Jonathan Eastoe, Grayson M Noah, Debargha Dutta, Alessandro Rossi, Jonathan D Fletcher, and Alberto Gomez-Saiz. 2024. Method for efficient large-scale cryogenic characterization of CMOS tech- nologies. IEEE Transactions on Instrumentation and Measurement (2024)

  15. [23]

    Xiang Fu, Michiel Adriaan Rol, Cornelis Christiaan Bultink, J Van Someren, Nader Khammassi, Imran Ashraf, RFL Vermeulen, JC De Sterke, WJ Vlothuizen, RN Schouten, et al. 2017. An experi- mental microarchitecture for a superconducting quantum processor. In Proceedings of the 50...

  16. [24]

    Gianluca Galletti, Fabian Paischer, Paul Setinek, William Hornsby, Lorenzo Zanisi, Naomi Carey, Stanislas Pamela, and Johannes Brand- stetter. 2025. 5D Neural Surrogates for Nonlinear Gyrokinetic Simu- lations of Plasma Turbulence. arXiv preprint arXiv:2502.07469 (2025)

  17. [25]

    Xujiao Gao, Erik Nielsen, Richard P Muller, Ralph W Young, An- drew G Salinger, Nathan C Bishop, Michael P Lilly, and Malcolm S Carroll. 2013. Quantum computer aided design simulation and opti- mization of semiconductor quantum dots. Journal of Applied Physics 114, 16 (2013)

  18. [26]

    Hubert C George, Mateusz T Madzik, Eric M Henry, Andrew J Wagner, Mohammad M Islam, Felix Borjans, Elliot J Connors, Joelle Corrigan, Matthew Curry, Michael K Harper, et al. 2024. 12-spin-qubit arrays fabricated on a 300 mm semiconductor manufacturing line. Nano Letters 25, 2 ...

  19. [27]

    Will Gilbert, Andre Saraiva, Wee Han Lim, Chih Hwan Yang, Arne Laucht, Benoit Bertrand, Nils Rambal, Louis Hutin, Christopher C Escott, Maud Vinet, et al. 2020. Single-electron operation of a silicon- CMOS 2× 2 quantum dot array with integrated charge sensing.Nano Letters 20, ...

  20. [28]

    Valentina Gualtieri, Charles Renshaw-Whitman, Vinicius Hernandes, and Eliska Greplova. 2025. QDsim: A user-friendly toolbox for sim- ulating large-scale quantum dot devices. SciPost Physics Codebases (2025), 046

  21. [29]

    Xingyi Guan, Joseph P Heindel, Taehee Ko, Chao Yang, and Teresa Head-Gordon. 2023. Using machine learning to go beyond poten- tial energy surface benchmarking for chemical reactivity. Nature Computational Science 3, 11 (2023), 965–974

  22. [30]

    Wonill Ha, Sieu D Ha, Maxwell D Choi, Yan Tang, Adele E Schmitz, Mark P Levendorf, Kangmu Lee, James M Chappell, Tower S Adams, Daniel R Hulbert, et al. 2021. A flexible design platform for Si/SiGe exchange-only qubits with low disorder. Nano Letters 22, 3 (2021), 1443–1448. 1...

  23. [31]

    Fabian Hader, Sarah Fleitmann, Jan Vogelbruch, Lotte Geck, and Stefan Van Waasen. 2024. Simulation of charge stability diagrams for automated tuning solutions (simcats). IEEE Transactions on Quantum Engineering (2024)

  24. [32]

    Ronald Hanson, Leo P Kouwenhoven, Jason R Petta, Seigo Tarucha, and Lieven MK Vandersypen. 2007. Spins in few-electron quantum dots. Reviews of modern physics 79, 4 (2007), 1217–1265

  25. [33]

    Will J Hardy, C Thomas Harris, Yi-Hsin Su, Yen Chuang, Jonathan Moussa, Leon N Maurer, Jiun-Yun Li, Tzu-Ming Lu, and Dwight R Luhman. 2019. Single and double hole quantum dots in strained Ge/SiGe quantum wells. Nanotechnology 30, 21 (2019), 215202

  26. [34]

    NW Hendrickx, DP Franke, A Sammak, G Scappucci, and M Veldhorst

  27. [35]

    Nico W Hendrickx, William IL Lawrie, Maximilian Russ, Floor Van Riggelen, Sander L De Snoo, Raymond N Schouten, Amir Sam- mak, Giordano Scappucci, and Menno Veldhorst. 2021. A four-qubit germanium quantum processor. Nature 591, 7851 (2021), 580–585

  28. [36]

    Toivo Hensgens, Takafumi Fujita, Laurens Janssen, Xiao Li, CJ Van Diepen, Christian Reichl, Werner Wegscheider, Sankar Das Sarma, and Lieven MK Vandersypen. 2017. Quantum simu- lation of a Fermi–Hubbard model using a semiconductor quantum dot array. Nature 548, 7665 (2017), 70–73

  29. [37]

    Joseph Hickie, Barnaby Van Straaten, Federico Fedele, Daniel Jirovec, Andrea Ballabio, Daniel Chrastina, Giovanni Isella, Georgios Kat- saros, and Natalia Ares. 2024. Automated long-range compensation of an rf quantum dot sensor. Physical Review Applied 22, 6 (2024), 064026

  30. [38]

    Yanxue Hong, AN Ramanayaka, Ryan Stein, MD Stewart, and JM Pomeroy. 2021. Developing single-layer metal-oxide-semiconductor quantum dots for diagnostic qubits. Journal of Vacuum Science & Technology B 39, 1 (2021)

  31. [39]

    Guangchong Hu, Wei Wister Huang, Ranran Cai, Lin Wang, Chih Hwan Yang, Gang Cao, Xiao Xue, Peihao Huang, and Yu He

  32. [40]

    Tsutomu Ikegami, Koichi Fukuda, Junichi Hattori, Hidehiro Asai, and Hiroyuki Ota. 2019. A TCAD device simulator for exotic mate- rials and its application to a negative-capacitance FET. Journal of Computational Electronics 18 (2019), 534–542

  33. [41]

    Wood, Jake Lishman, Julien Gacon, Simon Martiel, Paul D

    Ali Javadi-Abhari, Matthew Treinish, Kevin Krsulich, Christo- pher J. Wood, Jake Lishman, Julien Gacon, Simon Martiel, Paul D. Nation, Lev S. Bishop, Andrew W. Cross, Blake R. Johnson, and Jay M. Gambetta. 2024. Quantum computing with Qiskit. arXiv:2405.08810 [quant-ph] doi:10...

  34. [42]

    Debdeep Jena. 2022. Quantum physics of semiconductor materials and devices. Oxford University Press

  35. [43]

    Shui Jiang, Yi-Hua Chung, Chih-Chun Chang, Tsung-Yi Ho, and Tsung-Wei Huang. 2025. BQSim: GPU-accelerated Batch Quantum Circuit Simulation using Decision Diagram. InProceedings of the 30th ACM International Conference on Architectural Support for Program- ming Languages and Op...

  36. [44]

    J Robert Johansson, Paul D Nation, and Franco Nori. 2012. QuTiP: An open-source Python framework for the dynamics of open quantum systems. Computer physics communications 183, 8 (2012), 1760–1772

  37. [45]

    N Cody Jones, Rodney Van Meter, Austin G Fowler, Peter L McMahon, Jungsang Kim, Thaddeus D Ladd, and Yoshihisa Yamamoto. 2012. Layered architecture for quantum computing. Physical Review X 2, 3 (2012), 031007

  38. [46]

    Kalantre, Justyna P

    Sandesh S. Kalantre, Justyna P. Zwolak, Stephen Ragole, Xingyao Wu, Neil M. Zimmerman, M. D. Stewart, and Jacob M. Taylor. 2019. Machine learning techniques for state recognition and auto-tuning in quantum dots. npj Quantum Information 5, 1 (21 Jan 2019), 6

  39. [47]

    Mahmoud Khairy, Zhesheng Shen, Tor M Aamodt, and Timothy G Rogers. 2020. Accel-sim: An extensible simulation framework for validated gpu modeling. In 2020 ACM/IEEE 47th Annual International Symposium on Computer Architecture (ISCA) . IEEE, 473–486

  40. [48]

    Nader Khammassi, Randy W Morris, Shavindra Premaratne, Florian Luthi, Felix Borjans, Satoshi Suzuki, Robert Flory, Linda Patricia Os- una Ibarra, Lester Lampert, and Anne Y Matsuura. 2022. A scalable microarchitecture for efficient instruction-driven signal synthesis and coher...

  41. [49]

    Yoongu Kim, Weikun Yang, and Onur Mutlu. 2015. Ramulator: A fast and extensible DRAM simulator. IEEE Computer architecture letters 15, 1 (2015), 45–49

  42. [50]

    Leo P Kouwenhoven, Charles M Marcus, Paul L McEuen, Seigo Tarucha, Robert M Westervelt, and Ned S Wingreen. 1997. Elec- tron transport in quantum dots. Mesoscopic electron transport (1997), 105–214

  43. [51]

    Jan Krzywda, Weikun Liu, Evert van Nieuwenburg, and Oswin Krause. 2025. QDarts: A quantum dot array transition simulator for finding charge transitions in the presence of finite tunnel cou- plings, non-constant charging energies and sensor dots. SciPost Physics Codebases (2025), 043

  44. [52]

    Remi Lam, Alvaro Sanchez-Gonzalez, Matthew Willson, Peter Wirns- berger, Meire Fortunato, Ferran Alet, Suman Ravuri, Timo Ewalds, Zach Eaton-Rosen, Weihua Hu, et al. 2023. Learning skillful medium- range global weather forecasting.Science 382, 6677 (2023), 1416–1421

  45. [53]

    Maxime Lapointe-Major, Olivier Germain, Julien Camirand Lemyre, Dany Lachance-Quirion, Sophie Rochette, Félix Camirand Lemyre, and Michel Pioro-Ladrière. 2020. Algorithm for automated tuning of a quantum dot into the single-electron regime. Physical Review B 102, 8 (2020), 085301

  46. [54]

    Dylan H Liang, MengKe Feng, Philip Y Mai, Jesus D Cifuentes, An- drew S Dzurak, and Andre Saraiva. 2024. Electronic Correlations in Multielectron Silicon Quantum Dots. In 2024 IEEE 24th International Conference on Nanotechnology (NANO) . IEEE, 527–532

  47. [55]

    Junyi Liu, Yi Lee, Haowei Deng, Connor Clayton, Gengzhi Yang, and Xiaodi Wu. 2025. RISC-Q: A Generator for Real-Time Quantum Control System-on-Chips Compatible with RISC-V. arXiv preprint arXiv:2505.14902 (2025)

  48. [56]

    Jonathan Long, Evan Shelhamer, and Trevor Darrell. 2015. Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE conference on computer vision and pattern recognition . 3431–3440

  49. [57]

    Daniel Loss and David DiVincenzo. 1998. Quantum computation with quantum dots. Phys. Rev. A 57 (Jan 1998), 120–126. Issue 1. doi:10.1103/PhysRevA.57.120

  50. [58]

    Xiangyu Ma, Yuejing Wang, Joshua Zide, and Matthew Doty. 2020. Three-Electrode Device for Applying Two-Dimensional Vector Elec- tric Fields to Single In As Quantum Dots. Physical Review Applied 13, 6 (2020), 064029

  51. [59]

    Thomas McJunkin, ER MacQuarrie, Leah Tom, SF Neyens, JP Dodson, Brandur Thorgrimsson, J Corrigan, H Ekmel Ercan, DE Savage, MG Lagally, et al. 2021. Valley splittings in Si/SiGe quantum dots with a germanium spike in the silicon well. Physical Review B 104, 8 (2021), 085406

  52. [60]

    Thomas Walter McJunkin. 2021. Heterostructure modifications, fabri- cation improvements, and measurement automation of Si/SiGe quan- tum dots for quantum computation . The University of Wisconsin- Madison

  53. [61]

    Michael D McKay, Richard J Beckman, and William J Conover. 2000. A comparison of three methods for selecting values of input variables in the analysis of output from a computer code. Technometrics 42, 1 (2000), 55–61

  54. [62]

    Andrew J Miller, Will J Hardy, Dwight R Luhman, Mitchell Brick- son, Andrew Baczewski, Chia-You Liu, Jiun-Yun Li, Michael P Lilly, 13 Shize Che, Junyu Zhou, Seong Woo Oh, Jonathan Hess, Noah Johnson, Mridul Pushp, Robert Spivey, Anthony Sigillito, Gushu Li and Tzu-Ming Lu. 202...

  55. [63]

    AR Mills, DM Zajac, MJ Gullans, FJ Schupp, TM Hazard, and Jason R Petta. 2019. Shuttling a single charge across a one-dimensional array of silicon quantum dots. Nature communications 10, 1 (2019), 1063

  56. [64]

    Mills, Charles R

    Adam R. Mills, Charles R. Guinn, Michael J. Gullans, Anthony J. Sigillito, Mayer M. Feldman, Erik Nielsen, and Jason R. Petta. 2022. Two-qubit silicon quantum processor with operation fidelity exceed- ing 99%. Science Advances 8, 14 (2022), eabn5130

  57. [65]

    COMSOL Multiphysics. 1998. Introduction to comsol multiphysics ®. COMSOL Multiphysics, Burlington, MA, accessed Feb 9, 2018 (1998), 32

  58. [66]

    Samuel Neyens, Otto K Zietz, Thomas F Watson, Florian Luthi, Aditi Nethwewala, Hubert C George, Eric Henry, Mohammad Islam, An- drew J Wagner, Felix Borjans, et al. 2024. Probing single electrons across 300-mm spin qubit wafers. Nature 629, 8010 (2024), 80–85

  59. [67]

    Akito Noiri, Kenta Takeda, Takashi Nakajima, Takashi Kobayashi, Amir Sammak, Giordano Scappucci, and Seigo Tarucha. 2022. Fast universal quantum gate above the fault-tolerance threshold in silicon. Nature 601 (2022), 338–342

  60. [68]

    Stephan GJ Philips, Mateusz T Mądzik, Sergey V Amitonov, Sander L de Snoo, Maximilian Russ, Nima Kalhor, Christian Volk, William IL Lawrie, Delphine Brousse, Larysa Tryputen, et al. 2022. Universal control of a six-qubit quantum processor in silicon. Nature 609, 7929 (2022), 919–924

  61. [69]

    Anantha S Rao, Donovan Buterakos, Barnaby van Straaten, Valentin John, Cécile X Yu, Stefan D Oosterhout, Lucas Stehouwer, Giordano Scappucci, Menno Veldhorst, Francesco Borsoi, et al. 2025. Modular autonomous virtualization system for two-dimensional semiconduc- tor quantum do...

  62. [70]

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. 2015. U-net: Convolutional networks for biomedical image segmentation. In Med- ical image computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Munich, Germany, October 5-9, 2015, proceeding...

  63. [71]

    Daniel Sanchez and Christos Kozyrakis. 2013. ZSim: Fast and accu- rate microarchitectural simulation of thousand-core systems. ACM SIGARCH Computer architecture news 41, 3 (2013), 475–486

  64. [72]

    D Schröer, AD Greentree, L Gaudreau, K Eberl, LCL Hollenberg, JP Kotthaus, and S Ludwig. 2007. Electrostatically defined serial triple quantum dot charged with few electrons. Physical Review B—Condensed Matter and Materials Physics 76, 7 (2007), 075306

  65. [73]

    Andrii Sokolov, Dmytro Mishagli, Panagiotis Giounanlis, Imran Bashir, Dirk Leipold, Eugene Koskin, Robert Bogdan Staszewski, and Elena Blokhina. 2020. Simulation methodology for electron transfer in CMOS quantum dots. In Computational Science–ICCS 2020: 20th International Conf...

  66. [74]

    Alessio Spurio Mancini, Davide Piras, Justin Alsing, Benjamin Joachimi, and Michael P Hobson. 2022. CosmoPower: emulating cosmological power spectra for accelerated Bayesian inference from next-generation surveys. Monthly Notices of the Royal Astronomical Society 511, 2 (2022)...

  67. [75]

    Leandro Stefanazzi, Kenneth Treptow, Neal Wilcer, Chris Stoughton, Collin Bradford, Sho Uemura, Silvia Zorzetti, Salvatore Montella, Gustavo Cancelo, Sara Sussman, et al. 2022. The QICK (Quantum Instrumentation Control Kit): Readout and control for qubits and detectors. Review...

  68. [76]

    Michael Stein. 1987. Large sample properties of simulations using Latin hypercube sampling. Technometrics 29, 2 (1987), 143–151

  69. [77]

    Paul Steinacker, Tuomo Tanttu, Wee Han Lim, Nard Dumoulin Stuyck, MengKe Feng, Santiago Serrano, Ensar Vahapoglu, Rocky Y Su, Jonathan Y Huang, Cameron Jones, et al. 2025. Bell inequality vio- lation in gate-defined quantum dots. Nature communications 16, 1 (2025), 3606

  70. [78]

    Nard Dumoulin Stuyck, Andre Saraiva, Will Gilbert, Jesus Ci- fuentes Pardo, Ruoyu Li, Christopher C Escott, Kristiaan De Greve, Sorin Voinigescu, David J Reilly, and Andrew S Dzurak. 2024. CMOS compatibility of semiconductor spin qubits. arXiv preprint arXiv:2409.03993 (2024)

  71. [79]

    Masahiro Tadokoro, Takashi Nakajima, Takashi Kobayashi, Kenta Takeda, Akito Noiri, Kaito Tomari, Jun Yoneda, Seigo Tarucha, and Tetsuo Kodera. 2021. Designs for a two-dimensional Si quantum dot array with spin qubit addressability. Scientific Reports 11, 1 (2021), 19406

  72. [80]

    The CUDA-Q development team. [n. d.]. CUDA-Q. https://github. com/NVIDIA/cuda-quantum

  73. [81]

    I Thorvaldson, D Poulos, CM Moehle, SH Misha, H Edlbauer, J Reiner, H Geng, B Voisin, MT Jones, MB Donnelly, et al . 2025. Grover’s algorithm in a four-qubit silicon processor above the fault-tolerant threshold. Nature Nanotechnology (2025), 1–6

  74. [82]

    Florian K Unseld, Marcel Meyer, Mateusz T Mądzik, Francesco Borsoi, Sander L de Snoo, Sergey V Amitonov, Amir Sammak, Giordano Scappucci, Menno Veldhorst, and Lieven MK Vandersypen. 2023. A 2D quantum dot array in planar 28Si/SiGe. Applied Physics Letters 123, 8 (2023)

  75. [83]

    Wilfred G Van der Wiel, Silvano De Franceschi, Jeroen M Elzerman, Toshimasa Fujisawa, Seigo Tarucha, and Leo P Kouwenhoven. 2002. Electron transport through double quantum dots. Reviews of modern physics 75, 1 (2002), 1

  76. [84]

    Barnaby van Straaten, Federico Fedele, Florian Vigneau, Joseph Hickie, Daniel Jirovec, Andrea Ballabio, Daniel Chrastina, Giovanni Isella, Georgios Katsaros, and Natalia Ares. 2022. All rf-based tun- ing algorithm for quantum devices using machine learning. arXiv preprint arXi...

  77. [85]

    Barnaby van Straaten, Joseph Hickie, Lucas Schorling, Jonas Schuff, Federico Fedele, and Natalia Ares. 2024. QArray: A GPU-accelerated constant capacitance model simulator for large quantum dot arrays. SciPost Physics Codebases (2024), 035

  78. [86]

    LMK Vandersypen, H Bluhm, JS Clarke, AS Dzurak, R Ishihara, A Morello, DJ Reilly, LR Schreiber, and M Veldhorst. 2017. Interfacing spin qubits in quantum dots and donors—hot, dense, and coherent. npj Quantum Information 3, 1 (2017), 34

  79. [87]

    Lieven MK Vandersypen and Mark A Eriksson. 2019. Quantum computing with semiconductor spins. Physics Today 72, 8 (2019), 38–45

  80. [88]

    Menno Veldhorst, CH Yang, JCCea Hwang, W Huang, JP Dehollain, JT Muhonen, S Simmons, A Laucht, FE Hudson, Kohei M Itoh, et al

  81. [89]

    Thomas F Watson, SGJ Philips, Erika Kawakami, Daniel R Ward, Pasquale Scarlino, Menno Veldhorst, Donald E Savage, MG Lagally, Mark Friesen, Susan N Coppersmith, et al. 2018. A programmable two-qubit quantum processor in silicon. nature 555, 7698 (2018), 633–637

  82. [90]

    Aaron J Weinstein, Matthew D Reed, Aaron M Jones, Reed W An- drews, David Barnes, Jacob Z Blumoff, Larken E Euliss, Kevin Eng, Bryan H Fong, Sieu D Ha, et al. 2023. Universal logic with encoded spin qubits in silicon. Nature 615, 7954 (2023), 817–822

  83. [91]

    Yilun Xu, Gang Huang, Neelay Fruitwala, Abhi Rajagopala, Ravi K Naik, Kasra Nowrouzi, David I Santiago, and Irfan Siddiqi. 2023. QubiC 2.0: An extensible open-source qubit control system capa- ble of mid-circuit measurement and feed-forward. arXiv preprint arXiv:2309.10333 (2023)

  84. [92]

    Xiao Xue, Maximilian Russ, Nodar Samkharadze, Brennan Undseth, Amir Sammak, Giordano Scappucci, and Lieven M. K. Vandersypen

  85. [93]

    Yuchen Yang, Zhongtao Shen, Xing Zhu, Ziqi Wang, Gengyan Zhang, Jingwei Zhou, Xun Jiang, Chunqing Deng, and Shubin Liu. 2022. FPGA-based electronic system for the control and readout of super- conducting quantum processors. Review of Scientific Instruments 93, 7 (2022)

  86. [94]

    Jun Yoneda, Kenta Takeda, Tomohiro Otsuka, Takashi Nakajima, Matthieu R Delbecq, Giles Allison, Takumu Honda, Tetsuo Kodera, Shunri Oda, Yusuke Hoshi, et al. 2018. A quantum-dot spin qubit with coherence limited by charge noise and fidelity higher than 99.9%. Nature nanotechno...

  87. [95]

    David Zajac. 2018. Single Electron Spin Qubits in Sili- con Quantum Dots . Ph. D. Dissertation. Princeton Uni- versity. https://proxy.library.upenn.edu/login?url=https: //www.proquest.com/dissertations-theses/single-electron-spin- qubits-silicon-quantum-dots/docview/2128053747/se-2

  88. [96]

    DM Zajac, TM Hazard, Xiao Mi, E Nielsen, and Jason R Petta. 2016. Scalable gate architecture for a one-dimensional array of semicon- ductor spin qubits. Physical Review Applied 6, 5 (2016), 054013

  89. [97]

    DM Zajac, TM Hazard, X Mi, K Wang, and Jason R Petta. 2015. A reconfigurable gate architecture for Si/SiGe quantum dots. Applied Physics Letters 106, 22 (2015)

  90. [98]

    David M Zajac, Anthony J Sigillito, Maximilian Russ, Felix Borjans, Jacob M Taylor, Guido Burkard, and Jason R Petta. 2018. Resonantly driven CNOT gate for electron spins. Science 359, 6374 (2018), 439– 442

  91. [99]

    Xin Zhang, Elizaveta Morozova, Maximilian Rimbach-Russ, Daniel Jirovec, Tzu-Kan Hsiao, Pablo Cova Fariña, Chien-An Wang, Stefan D Oosterhout, Amir Sammak, Giordano Scappucci, et al. 2025. Universal control of four singlet–triplet qubits. Nature Nanotechnology 20, 2 (2025), 209–215

  92. [100]

    Joshua Ziegler, Florian Luthi, Mick Ramsey, Felix Borjans, Guoji Zheng, and Justyna P Zwolak. 2023. Automated extraction of capaci- tive coupling for quantum dot systems. Physical Review Applied 19, 5 (2023), 054077

  93. [101]

    Justyna P Zwolak, Thomas McJunkin, Sandesh S Kalantre, JP Dodson, ER MacQuarrie, DE Savage, MG Lagally, SN Coppersmith, Mark A Eriksson, and Jacob M Taylor. 2020. Autotuning of double-dot devices in situ with machine learning. Physical review applied 13, 3 (2020), 034075

  94. [102]

    Justyna P Zwolak and Jacob M Taylor. 2023. Colloquium: Advances in automation of quantum dot devices control. Reviews of modern physics 95, 1 (2023), 011006

  95. [103]

    #converged

    Justyna P Zwolak, Jacob M Taylor, Reed W Andrews, Jared Benson, Garnett W Bryant, Donovan Buterakos, Anasua Chatterjee, Sankar Das Sarma, Mark A Eriksson, Eliška Greplová, et al. 2024. Data needs and challenges for quantum dot devices automation. npj Quantum Information 10, 1 ...

  96. [2015]

    Nature 526, 7573 (2015), 410–414

    A two-qubit logic gate in silicon. Nature 526, 7573 (2015), 410–414

  97. [2020]

    Nature 577, 7791 (2020), 487–491

    Fast two-qubit logic with holes in germanium. Nature 577, 7791 (2020), 487–491

  98. [2022]

    Nature 601 (2022), 343–347

    Quantum logic with spin qubits crossing the surface code threshold. Nature 601 (2022), 343–347. 14 NeuroQD: A Learning-Based Simulation Framework For Quantum Dot Devices

  99. [2025]

    Intelligent Computing 4 (2025), 0115

    Single-Electron Spin Qubits in Silicon for Quantum Computing. Intelligent Computing 4 (2025), 0115

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.