This is a personal working roadmap, distilled from my current notes. The point is not to claim that one person can build a universal quantum computer. The real goal is more grounded: learn quantum computation seriously, use real cloud hardware, and eventually build a small apparatus that can manipulate a quantum degree of freedom in a reproducible way.
这是一个个人 working note，整理自我目前的两份量子计算自学与 DIY 路线笔记。这里的目标不是宣称 一个人可以造出通用量子计算机，而是更实际地：认真学习量子计算，使用云端真机，并尝试搭建一套 低成本、可复现、能真实操纵量子自由度的小型实验装置。
The two longer source notes now live as separate subpages: an eight-week Step 0-1 syllabus for software and theory, and a DIY hardware roadmap for tabletop optics, low-field NMR, and research-group-scale options.
两份更详细的源笔记已经整理成独立子页面： 八周 Step 0-1 自学清单 负责软件与理论部分； DIY 硬件路线图 负责桌面光学、低场 NMR 和课题组量级硬件选项。
What “build” means here这里的“造”是什么意思
In this note, "building a quantum computer" means building toward experimental control of a qubit-like or quantum-coherent system. A cloud quantum program is real quantum computing but not my own hardware. A tabletop optics or low-field NMR setup is not a scalable universal machine, but it can be a real experimental bridge between mathematics, physics, and engineering.
在这份笔记里，“造一台量子计算机”的含义是：逐步走向对类 qubit 或量子相干系统的实验控制。 云端量子机是真正的量子计算，但不是自己的硬件；桌面光学或低场 NMR 装置并不是可扩展的通用机器， 但可以成为数学、物理和工程之间真实可动手的桥梁。
- Be honest about scale: IBM/Google-style systems need industrial and laboratory infrastructure.诚实面对尺度：IBM/Google 级别系统需要工业与实验室基础设施。
- Start where feedback is fastest: simulation, Qiskit, and cloud hardware.从反馈最快的地方开始：模拟器、Qiskit 和云端真机。
- Move to hardware only when the gates, measurement, noise, and calibration ideas are no longer abstract.只有当量子门、测量、噪声和校准不再只是抽象概念时，再进入硬件。
Roadmap路线图
A. Cloud quantum hardware and simulatorsA. 云端量子机与模拟器
This should be step zero. I want to use Qiskit, local simulators, and IBM Quantum hardware to build intuition for states, gates, circuits, measurement, noise, and hybrid algorithms before touching any optical table or circuit board. I keep the full weekly plan on the Step 0-1 syllabus page.
这是第零步。先用 Qiskit、本地模拟器和 IBM Quantum 云端真机建立对量子态、量子门、电路、测量、 噪声和混合算法的手感，再考虑光学平台或电路板。完整周计划放在 Step 0-1 自学清单里。
Milestones: Bell states, GHZ states, Deutsch-Jozsa, small Grover experiments, phase estimation, and a first VQE notebook. 里程碑：Bell 态、GHZ 态、Deutsch-Jozsa、小规模 Grover、相位估计，以及第一个 VQE notebook。
B. Tabletop polarization opticsB. 桌面偏振光学装置
A realistic first hardware attempt is a tabletop optics setup where photon polarization plays the role of a qubit. Polarizers, wave plates, beam splitters, and photodetectors can demonstrate preparation, single-qubit gates, and measurement. With better equipment, this path can move toward single-photon experiments and Bell-type demonstrations. The broader hardware comparison is on the DIY hardware roadmap page.
第一个硬件尝试可以是桌面偏振光学：用光子偏振作为 qubit 的物理载体，用偏振片、波片、分束器和探测器 演示态制备、单比特门和测量。设备更好时，可以进一步做单光子实验和 Bell 型演示。更完整的硬件比较放在 DIY 硬件路线图里。
Possible outputs: a reproducible teaching setup, a short demo video, and a written lab note explaining every optical component as a quantum operation. 可能产出：一套可复现的教学装置、一个演示视频，以及一份把每个光学元件对应到量子操作的实验笔记。
C. Earth’s-field NMRC. 地磁场 NMR
Another route is a low-field NMR-style device. It is not a standard universal quantum computer, but it can show spin control, spin echo, nutation, and coherent readout. This route forces me to learn electronics, signal processing, and the Bloch-equation side of quantum control.
另一条路线是低场 NMR 装置。它不是标准意义上的通用量子计算机，但可以展示自旋控制、自旋回波、 章动和相干读出。这条路会逼迫我学习电子学、信号处理和 Bloch 方程意义下的量子控制。
Possible outputs: a PCB or microcontroller-based prototype, spin-echo data, calibration plots, and a public build log. 可能产出：一个基于 PCB 或微控制器的原型、自旋回波数据、校准图和公开 build log。
D. Research-group scale hardwareD. 课题组量级硬件
Photonic gates, trapped ions, superconducting qubits, and scalable architectures belong to a different scale: stable laboratories, experienced supervisors, vacuum or cryogenic infrastructure, lasers or microwave engineering, and years of iteration. For me, this is a possible graduate-school direction rather than a solo weekend project.
光子量子门、离子阱、超导量子比特以及可扩展架构属于另一种尺度：稳定实验室、有经验的导师、 真空或低温基础设施、激光或微波工程，以及多年迭代。对我来说，这更像是未来研究生阶段可能进入的方向， 不是一个人周末就能完成的项目。
First eight weeks前八周学习计划
The first phase is software and theory. The goal after two months is to run small circuits on real cloud hardware and understand the physics behind each line of code.
第一阶段先做软件和理论。两个月后的目标是：能够在云端真机上跑小电路，并理解每一行代码背后的物理含义。
| Weeks周次 | Focus重点 | Deliverable产出 |
|---|---|---|
| 1-2 | Dirac notation, Hilbert spaces, one-qubit gates, Bloch sphere, Qiskit setup.Dirac 记号、Hilbert 空间、单比特门、Bloch 球、Qiskit 环境。 | Single-qubit notebooks and a gate cheat sheet.单比特 notebook 和量子门速查表。 |
| 3-4 | Tensor products, Bell states, GHZ states, controlled gates, first cloud run.张量积、Bell 态、GHZ 态、受控门、第一次云端真机运行。 | Bell/GHZ circuits with simulator and hardware comparisons.Bell/GHZ 电路，并比较模拟器和真机结果。 |
| 5-6 | Density matrices, noise channels, Deutsch-Jozsa and Bernstein-Vazirani.密度矩阵、噪声通道、Deutsch-Jozsa 和 Bernstein-Vazirani。 | A short note on noise and a query-algorithm notebook.一篇噪声笔记和一个查询算法 notebook。 |
| 7-8 | Grover search, QFT, phase estimation, and a small final project.Grover 搜索、QFT、相位估计和一个小结项项目。 | A GitHub repo with README, notebooks, plots, and a short reflection.一个包含 README、notebook、图和短反思的 GitHub 仓库。 |
Principles原则
- Every theoretical concept should end in a circuit, a plot, or an experiment.每个理论概念最好都落到一个电路、一张图或一个实验。
- Every hardware component should have a mathematical interpretation.每个硬件元件都应该有对应的数学解释。
- Keep the first device small enough to finish and honest enough to explain.第一台装置要足够小，能做完；也要足够诚实，能讲清楚。
- Treat budget, vendors, and cloud access as time-sensitive details that must be rechecked before buying anything.预算、供应商和云端权限都可能变化，真正购买或注册前必须重新确认。
References参考资料
- IBM Quantum Documentation
- IBM Quantum Learning
- Install Qiskit
- DIY budget quantum computer, Hackaday.io
- Thorlabs Quantum Optics Educational Kits
- Nielsen and Chuang, Quantum Computation and Quantum Information.
- Preskill, Lecture Notes on Quantum Computation.