Quantum Dot Control: New Pulses Boost Accuracy for Future Computers

Julian D. Teske and colleagues at Q-CTRL and Intel have shown that designing pulses with near-zero time integrals effectively cancels distortions that commonly limit the accuracy of quantum gate operations. The approach achieves comparable gate fidelities to conventional filtering methods, but with a sharply reduced need for calibration parameters. This simplification offered by these “quasi-zero” pulses represents a key step towards scalable and automated calibration procedures for future quantum computers. Reduced calibration parameters achieve high-fidelity qubit control using quasi-zero pulses Error rates in single-qubit gates on the Intel Tunnel Falls six-dot device fell below 1%, matching the performance of traditional filtering techniques, despite employing fewer calibration parameters. Achieving such fidelity previously demanded extensive calibration of numerous parameters to counteract pulse distortions, and complex filtering was essential for accurate qubit control. These distortions arise from the inherent limitations of generating and delivering precise electrical signals to nanoscale quantum dots, where the qubits are physically realised. The transfer function describing these distortions is often complex and requires characterisation of many parameters, typically involving a detailed mapping of the control signal’s evolution as it propagates to the qubits. By generalizing ‘net-zero’ pulse designs into ‘quasi-zero’ pulses, a controlled residual signal was enabled, streamlining the calibration process and paving the way for scalable automation in future quantum processors. Net-zero pulses, while effective, offer limited flexibility; quasi-zero pulses introduce a carefully calibrated offset, providing a balance between distortion cancellation and control authority. This simplification is particularly advantageous for larger systems where calibration complexity rapidly increases, offering a pathway towards practical, large-scale quantum computation. The calibration problem scales non-linearly with the number of qubits, meaning that doubling the qubit count can more than double the calibration effort. The new ‘quasi-zero’ pulse designs, when applied to the Intel Tunnel Falls six-dot device, maintained fidelity while sharply reducing calibration needs; the system achieved performance comparable to traditional filtering techniques with fewer tunable parameters. Detailed analysis revealed these optimised pulses exhibit a duration identical to previously filtered pulses, ensuring no trade-off in speed during gate operations. Maintaining pulse duration is crucial, as longer pulses introduce decoherence, a process where quantum information is lost due to interactions with the environment.

The Intel Tunnel Falls device utilises silicon quantum dots fabricated using advanced semiconductor manufacturing techniques, allowing for precise control over the electron confinement and qubit properties. Complete gate sets for exchange-only qubits were successfully benchmarked through both simulations and physical experimentation on the device, confirming the strong durability of the approach. Exchange-only qubits leverage the exchange interaction between two electrons confined in adjacent quantum dots to mediate qubit coupling and implement two-qubit gates. This approach avoids the need for direct control of charge states, simplifying the control scheme and reducing sensitivity to charge noise. The benchmarking process involved characterising the performance of various single- and two-qubit gates, including rotations and entangling operations, to verify the overall functionality of the qubit system. This advancement builds upon earlier work in silicon spin qubits, including the fabrication of 12-qubit arrays on 300mm wafers, and extends the possibilities for automated calibration schemes vital for scaling quantum processors. The use of 300mm wafers is significant, as it aligns with standard semiconductor manufacturing practices, potentially enabling cost-effective mass production of quantum processors. Addressing calibration bottlenecks enables progression towards scalable quantum processors The authors acknowledge that their current demonstration relies on a relatively small six-qubit device, despite simplifying calibration being a clear win for building larger quantum processors. Scaling this ‘quasi-zero’ pulse technique to significantly larger and more complex quantum dot arrays presents a substantial, and as yet unaddressed, challenge.

The team notes that charge configuration at negative pulse values, and potential crosstalk between qubits, could become more problematic as qubit numbers increase, potentially eroding the gains made in calibration speed. Crosstalk occurs when the control signal intended for one qubit inadvertently affects neighbouring qubits, leading to errors. Mitigating crosstalk requires careful design of the qubit layout and control circuitry, as well as advanced calibration techniques to compensate for the unwanted interactions. Furthermore, variations in device fabrication can lead to inconsistencies in qubit properties, necessitating individual calibration of each qubit in a large array. Even acknowledging the challenges of scaling this technique to larger qubit numbers and managing potential interference, this advance represents a major step forward for quantum computing. The process of fine-tuning quantum processors for accuracy, reducing the complexity of calibration, is vital for building practical devices. This ‘quasi-zero’ pulse design demonstrably simplifies calibration without sacrificing performance on Intel’s six-qubit test chip, paving the way for faster automation and, ultimately, more powerful quantum computers. Automated calibration is essential for managing the increasing complexity of large-scale quantum processors, reducing the need for manual intervention and enabling continuous optimisation of qubit performance. The ability to rapidly calibrate and optimise a quantum processor is crucial for maintaining its accuracy and reliability over time. Calibration of quantum processors has been simplified by using ‘quasi-zero’ pulses, reducing the need for complex tuning. This new approach maintains performance on existing six-qubit devices, offering a pathway to automated control and larger systems. Quasi-zero pulses represent a major advancement in controlling spin qubits, the fundamental building blocks of quantum computers, utilising the spin of an electron confined within nanoscale ‘quantum dots’. These quantum dots, acting as artificial atoms, provide a solid-state platform for realising qubits with long coherence times and strong interactions. This technique generalises existing ‘net-zero’ pulse designs, allowing for a small residual signal which streamlines the calibration process, a critical step in maintaining accuracy. By reducing the number of adjustable parameters needed for precise control, comparable gate fidelities to conventional filtering methods on Intel’s Tunnel Falls six-dot device have been demonstrated. The achievement of gate fidelities below 1% is a significant milestone, bringing silicon spin qubits closer to the threshold required for fault-tolerant quantum computation. The research successfully demonstrated a new pulse design, termed ‘quasi-zero’, for controlling spin qubits in quantum dots. This method simplifies the calibration of quantum processors by reducing the number of adjustable parameters needed to achieve high gate fidelities, comparable to those obtained with more complex filtering techniques on a six-qubit device. Maintaining performance with fewer tuning parameters is important because it enables faster automation of the calibration process. The authors suggest this reduction in complexity is a step towards building and optimising larger-scale quantum devices. 👉 More information 🗞 Driving Exchange Interaction in Spin Qubits with Quasi-Zero Pulses 🧠 ArXiv: https://arxiv.org/abs/2606.07472 Stay current. See today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals. Tags: