Tetsuo Kodera

Cross-correlated charge noise in silicon quantum dots has emerged as an important issue for fault-tolerant quantum computation, as it has been suggested to both limit quantum error correction performance and reveal unique information about noise sources. We develop a method to characterize charge noise cross-correlations in a semiconductor qubit device based solely on quantum-dot transport current. By dynamically switching the site-selective sensitivity of transport current through a double quantum dot to charge noise, we extract potential fluctuations at individual dots without requiring additional sensing devices or spin operations. Based on the obtained power spectral densities, we discuss the coupling of a dominant two-level fluctuator to individual dots. We furthermore verify the consistency of the measurement protocol by applying artificial noise via gate electrodes and investigate the temperature dependence of charge noise correlations. The method offers a convenient and versatile alternative approach that substantially facilitates the assessment of charge noise correlations in diverse semiconductor qubit devices.

Semiconductor spin qubits are compatible with CMOS processes and support high integration density, making them promising for scalable quantum computing. Yet multiplexed readout remains a bottleneck. Frequency multiplexing is bandwidth-limited, while time multiplexing introduces sequential latency and extra power. We adopt in-phase ( I )–quadrature ( Q ) encoding, mapping readout signals to the complex ( I – Q ) plane to enable parallel multi-qubit discrimination and compact, digitally processable representations, compatible with both resonator-based and inductor-less electrical readout. We develop machine-learning classifiers trained on synthesized IQ datasets that emulate spin-qubit measurements with noise, drift, and varied constellation layouts. Across wide signal-to-noise ratios, the models outperform conventional thresholding and sustain high accuracy, demonstrating robust state discrimination under realistic conditions. By improving qubit-state decoding in a digitally scalable form, this work supports larger spin-qubit arrays and more efficient cryogenic readout architectures.

Statistical evaluation of two-level fluctuators (TLFs) in quantum dots is essential for improving the coherence of silicon spin qubits. However, conventional methods detect TLFs only under high-sensitivity conditions, which limits the accessible voltage range for characterization. Here we introduce a method that detects individual switching of TLFs in a silicon quantum dot even at low-sensitivity conditions. This method can extract the voltage dependence near the conditions activating the TLF switching, which provides insight into TLF switching mechanisms. Our approach facilitates routine statistical characterization of noise caused by TLFs in silicon spin qubits.

Spin qubits in silicon quantum dots (QDs) are promising candidates for scalable implementation, but increasing qubit numbers requires higher wiring density and reduced signal integrity. This study addresses these interconnect challenges by employing an active silicon interposer as an alternative to conventional cryogenic CMOS implementations. We directly implement CMOS selector circuits on the interposer to improve wiring efficiency and enable flexible selection among multiple QDs. We evaluate their DC characteristics, switching response, and pulse bandwidth under cryogenic conditions, confirming stable operation and precise QD control. The selector achieves switching rate up to 1 kHz and pulse transmission up to 20 MHz, demonstrating effective control within the low-frequency regime, where operations like DC characterization of QDs and baseband pulse for QD potential control are typically conducted. These results show that integrating active circuitry on a silicon interposer can reduce wiring complexity and improve measurement efficiency in multi-qubit systems.

We report an improvement in the readout accuracy of a silicon spin qubit using signal processing based on log-likelihood ratio (LLR). This technique is a statistical method that enhances the detection accuracy of weak signals with relatively low computational cost. We apply this technique to spin qubit readouts, where a qubit state is encoded to the presence or absence of a pulse-like signal. Our experimental results demonstrate that the LLR technique enables reliable detection of such pulse-like signals even when a wide readout bandwidth reduces the signal-to-noise ratio below one. We also confirm the performance from simulations and find an improvement in detection accuracy consistent with the experimental results. We believe that this result contributes to the realization of fast and frequent qubit readout within the coherence time for quantum error correction.

Surface acoustic wave (SAW) technology has been explored in thin-film materials to discover fundamental phenomena and to investigate their physical properties. It is used to excite and manipulate quasiparticles such as phonons or magnons and can dynamically modulate the properties of the materials. In the field, SAWs are typically excited by a continuous wave at a resonant frequency. Recently, generation of a single-cycle SAW pulse has been demonstrated on GaAs substrate. Such a SAW pulse provides a potential to access a single quasiparticle excitation and to investigate its dynamics by time-resolved measurements. On the other hand, to modulate and control the properties of thin-film materials, it is generally required to generate high-intensity SAWs. In this work, we demonstrate the efficient generation of a SAW pulse using a chirp interdigital transducer (IDT) on LiNbO3 substrate. We have fabricated chirp IDT devices with bandwidths from 0.5 GHz to 5.5 GHz. We also confirmed the generation of a SAW pulse with 0.3 ns full width at half maximum by performing time-resolved measurements. The conversion efficiency between input power and SAW on LiNbO3 substrate is approximately 45 times larger than that on GaAs substrate. This enables us to generate a high-intensity SAW pulse, meeting the requirement for the modulation of thin films. Our results will expand the research in the field, such as spintronics and magnonics, and lead to their further advancements.

Cryogenic control systems are essential for integrating semiconductor spin qubits, and reducing phase noise in phase-locked loop (PLL) frequency synthesizers is critical for high-fidelity spin qubit control. Conventional single-frequency phase noise evaluation methods are not suitable for the multi-frequency demands of spin qubit architectures. To address this, we develop a two-dimensional (2D) phase noise mapping method that evaluates spectral purity as a function of both output and offset frequency. Using a cryogenic fractional-N PLL, we measure phase noise under different reference frequencies and observe frequency dependence of spurious signal. Simulations based on these measurements show infidelity variations up to ∼0.08% across 19.60–20.40 GHz. Our method enables identification of optimal signal configurations and supports resource-efficient design of control electronics. This framework contributes to a practical tool for scaling spin qubit systems and improving signal quality in cryogenic quantum processors.

Semiconductor quantum dots (QDs) are promising hosts for quantum computers because of their scalability. In order to expedite the development process, there is a strong need for fully automated tuning of QDs that allows for en masse characterization of newly fabricated devices and control over large-scale systems with appreciable variability. Machine learning has been actively explored as a means to this end; however, challenges remain in terms of versatility for different tasks and device types. In this study, we explore a model-based reinforcement learning (MBRL) approach: unlike traditional reinforcement learning techniques, the learning process of MBRL progresses by constructing a model for the environment, which is to be diverted for other tasks and/or devices, thereby minimizing time-consuming learning processes. Using pre-measured data, we construct an environment model and, despite the intrinsic sparse reward distribution of the QD system, demonstrate its suitability for MBRL by emulating the process of auto-tuning to a single QD region. Our results highlight the potential of MBRL for more generic QD auto-tuning techniques, providing a promising step toward fully automated QD tuning.

Frequency-multiplexed radio-frequency (RF) reflectometry is a promising technique for large-scale quantum dot-based qubit systems because it enables simultaneous readout of multiple qubits and helps to reduce the wiring density in peripheral circuits. However, in the conventional L-C-type matching circuits, resonator frequency tunability is limited by poor design flexibility of the shunt capacitance and hence the circuit. Recently, L-L-type matching circuits comprising a series and a shunt inductor have been proposed to address this issue. Here, we report RF reflectometry readout of silicon quantum dot-based charge sensors embedded in L-L-type circuits. We determine the inductance values for target resonance frequencies based on our equivalent circuit model that incorporates the parasitic elements of the inductors and demonstrate resonator frequency tunability of up to 1 GHz with 3-dB resonator bandwidths on the order of MHz. The measured conductance sensitivities, on the order of 10−5(e2/h)/Hz at 4 K, are only moderately dependent on the resonance frequency up to 630 MHz. These results highlight the prospects of L-L-type matching circuits for rapid and multiplex reflectometry readout of semiconductor spin qubits.