Scalable and Automated Tuning of Spin-based Quantum Computer Architectures

基于自旋的量子计算机体系结构的可扩展和自动调谐简要描述了研究的背景，包括潜在的影响：-量子计算是一种新兴技术，有可能使用新的计算范式解决经典的棘手问题。这种计算的基础是使用系统的量子态来编码信息，这与现代晶体管（比特）的二进制状态不同。在实现这些量子比特（qubit）的许多候选平台中，限制在半导体结构中的自旋具有吸引力，因为它们的相干时间长，尺寸小，易于与控制电子设备集成，使它们在可扩展性方面具有优势。挑战在于从概念验证设备（仅包含少数量子位）转向包含数十个此类量子位甚至更多的微芯片。这目前受到与主体材料相关联的设备可变性和噪声的阻碍，在第一种情况下，这使得设备调谐非常困难，并且在第二种情况下，可能导致量子操作的保真度降低，特别是如果这些噪声源具有时空相关性。如果能够设计出聪明的算法和协议，首先有效地调谐多个设备，其次实时补偿噪声，这将标志着自旋量子位架构的可扩展性向前迈出了巨大的一步。目的和目标：该项目的目的是探索能够在大长度尺度上调整和稳定自旋量子比特的机器学习方法。通过这样做，希望能够激发自旋量子位架构的发展，这些架构对可变性和噪声最有弹性。到目前为止，自旋量子位架构一直是以物理思维为基础的，即自旋量子位如何最容易相互作用，以及如何使用磁场和电场最容易地解决它们。该项目方法的新奇在于软件和硬件的协同设计，这在以前的设备设计阶段没有考虑过。这将是一项跨学科的工作，利用实验物理学和软件方法，例如加速贝叶斯优化方法和时间序列分析-这是基于自旋的量子计算中的一项新工作。该项目的一些里程碑将包括晶圆级材料无序的表征，将调谐算法扩展到少数量子位之外，以及补偿噪声的新方法，这将通过增强量子操作的保真度来证明。可以研究的一些材料平台包括Ge/SiGe异质结构、Si-MOS和Si Fin-FET器件，目的是隔离最有前途的候选材料。EPSRC校准：-该项目福尔斯EPSRC量子技术研究领域，任何参与的公司或合作者：-与奥地利科学技术研究所（ISTA）Katsaros集团的现有合作将得到扩展，我们将从他们那里获得量子设备，我们的算法和实验将在其上进行测试。此外，设备可以来自巴塞尔大学的量子相干实验室或位于苏黎世的IBM研究实验室的量子技术部门或QuTech的Scappucci实验室。此外，为了在晶圆级上表征量子器件，我们将与IBM或格勒诺布尔的CEA-Leti团队联系，使用他们最先进的低温探测器系统。最后，我们将继续与牛津大学工程科学系和统计学系的团队合作，他们提供了宝贵的算法输入。

Scalable and Automated Tuning of Spin-based Quantum Computer ArchitecturesBrief description of the context of the research including potential impact: - Quantum computing is an emergent technology that has the potential to solve classically intractable problems using a new paradigm of computation. The basis for such computation uses the quantum states of a system to encode information, unlike the binary states of modern-day transistors (bits). Of the many candidate platforms for realising these quantum bits (qubits), spins confined in semiconductor structures are attractive for their long coherence times as well as small size and ease of integration with control electronics, giving them an edge in terms of scalability. The challenge lies in moving away from proof-of-concept devices (containing just a handful of qubits) to microchips that contain dozens of such qubits and beyond. This is currently hindered by device variability and noise associated with the host material, which, in the first case, makes device tuning very difficult, and in the second case, can lead to a reduction in the fidelity of quantum operations, especially if these noise sources have spatio-temporal correlations. If clever algorithms and protocols can be devised to, firstly, tune multiple devices efficiently, and secondly, compensate for noise in real-time, this would mark an enormous step forward in the scalability of spin qubit architectures. Aims and objectives:- The aim of this project is to explore machine learning methods capable of tuning and stabilizing spin qubits over large length-scales. In doing so, the hope is to inspire the development of spin qubit architectures that are most resilient to variability and noise. Until now, spin qubit architectures have been pioneered with a physics mindset, i.e. how spin qubits can most easily interact with one another and how they can be most easily addressed using magnetic and electric fields. The novelty of this project approach lies in the software and hardware co-design, which has previously not been considered at the device design stage. This will be an interdisciplinary effort, drawing on both experimental physics and software methodologies, such as accelerated Bayesian optimization methods and time series analysis - a new effort within spin-based quantum computing. Some milestones for this project would include the characterization of material disorder at the wafer scale, expansion of tuning algorithms beyond a handful of qubits, and new methods for compensating noise, as would be evidenced by an enhancement in the fidelity of quantum operations. Some material platforms that could be investigated include Ge/SiGe heterostructures, Si-MOS, and Si Fin-FET devices, with the goal to isolate the most promising candidate. EPSRC alignment:- This project falls within the EPSRC Quantum Technologies research area.Any companies or collaborators involved:- An existing collaboration with Katsaros Group at the Institute of Science and Technology Austria (ISTA) will be extended, from whom we will receive quantum devices on which our algorithms and experiments will be tested. In addition, devices could be sourced from the Quantum Coherence Laboratory at the University of Basel or the Quantum Technologies division at the IBM Research Laboratory in Zurich or Scappucci Lab at QuTech. Moreover, to characterize quantum devices on a wafer scale, we will reach out to the teams at IBM or at CEA-Leti in Grenoble to use their state-of-the-art cryo-prober system. Lastly, we will continue collaborations with groups at the University of Oxford from departments of Engineering Science as well as Statistics, who provide invaluable algorithmic input.