Scanning Probe Fabrication and Readout of Atomically Precise Silicon Quantum Technologies

量子计算有望在金融、医学、密码学和材料模拟等许多应用领域取得巨大进步。量子计算机的基本计算元素是量子比特或量子位，要执行真正具有变革性的计算，需要大量的量子位（可能有数百万）。我们建议开发一种工艺，以制造比以前更多的量子位。此外，我们将使用材料系统硅来实现这一目标，这与目前唯一可用于大规模市场生产多量子位量子计算机的现成工业直接兼容。我们的量子比特将由硅中的杂质原子构成，也就是掺杂剂。这是用扫描隧道显微镜（STM）完成的，它用一个非常锋利的探针尖“感觉”表面上的原子，就像录音机感觉黑胶唱片的凹槽一样。在此之前，只有一对掺杂量子位被制造出来，而且没有办法放大到一个有用的数字。为了移动和观察组成量子计算机的数百万个掺杂原子，我们需要先进的机器控制和数据处理工具。纳米层研究计算将使用他们首创的专有机器学习软件来训练STM在没有人为干预的情况下自行执行控制和处理任务。使用这种类型的数据处理，也被称为人工智能（AI），特别适合于图像处理和模式识别，并且可以用来找到众所周知的“大海捞针”。当我们使用扫描隧道显微镜移动并观察硅量子计算机组件中的大量原子时，这正是我们需要做的。简而言之，该项目利用人工智能控制原子分辨率显微镜，精确定位硅中大量单个杂质原子。这项技术将使最终制造出硅基量子计算机成为可能。

Quantum computing promises tremendous advances in a number of applications, including finance, medicine, cryptography, and materials simulation. The fundamental computation element of a quantum computer is the quantum bit, or qubit, and to perform calculations that are truly transformative, huge numbers of qubits (potentially millions) are required. We propose to develop a process for fabricating many more qubits than have previously been made. Furthermore, we will do so using the material system silicon, which is directly compatible with the only ready-made industry currently available for largescale market production of a many qubit quantum computer.Our qubits will be made from impurity atoms in silicon, known as dopants. This is done using a scanning tunnelling microscope (STM), which "feels" the atoms on a surface with an extremely sharp probe tip, much like an audio record player feels the grooves of a vinyl record. Previously, only a pair of dopant qubits have been made, and with no route to scaling-up to a useful number. In order to move and see the millions of individual dopant atoms that will make up a quantum computer, we require advanced machine controls and data processing tools. Nanolayers Research Computing will use the proprietary machine learning software they have pioneered to train the STM to perform control and processing tasks on its own, without human user intervention. The use of this type of data processing, also known as artificial intelligence (AI), is particularly well-suited to image processing and pattern recognition, and can be used to find the proverbial "needle in a haystack". This is exactly what needs to be done when we use a scanning tunnelling microscope to move and then see a large collection of atoms in a silicon quantum computer component. In short, this project uses artificial intelligence to control an atomic resolution microscope and precisely position a large number of individual impurity atoms in silicon. This technology will enable the eventual fabrication of a silicon-based quantum computer.