Skip to main content

Learning path

Quantum hardware and error correction

“I want to work on building quantum computers that actually work.”

Graduate students, lab engineers and researchers moving into quantum hardware, control or quantum error correction.

4 milestones · 14 steps · 32 minutes of reading, plus hands-on time

By the end, you will be able to

  • Characterise noise: decoherence, readout error, and how to measure them
  • Explain the surface code and why a threshold exists
  • Train and compare decoders on your own noise model
  • Calibrate, mitigate, and run a reproducible experiment on real hardware

 

Start: Decoherence

  1. Milestone 1: What goes wrong

    4 steps

    You can name and model every way a physical qubit loses information.

  2. Milestone 2: Error correction

    4 steps

    You understand how logical qubits are built out of bad physical ones.

    • LessonAdvanced20 min
      Quantum Error Correction

      Syndromes, stabilisers, and the threshold theorem.

      Auto-tracks
    • VisualizationAdvanced
      The Surface Code

      The leading candidate code, on a lattice you can inject errors into.

    • VisualizationAdvanced
      Decoder Duel: Lookup vs Learned

      Lookup table versus learned decoder, head to head.

    • Workbench
      Decoder Lab

      Train a decoder and watch the threshold behaviour emerge.

  3. Milestone 3: Working with a real device

    4 steps

    You can tune, mitigate and route for an actual chip.

  4. Milestone 4: Reproducible research

    2 steps

    You can reproduce a published result and record your own well enough for others to.

    • Workbench
      Run This Paper

      Landmark papers, executed live — check the claims yourself.

    • Workbench
      Experiment Runner

      An experiment record an agent can run and a colleague can read.

Progress is stored in this browser only. Lessons tick themselves when you finish them and challenges when you solve them; everything else you tick yourself, because opening a page is not the same as understanding it.

After this path

  • Reproduce a landmark result in Run This Paper and publish the run permalink alongside your notes.
  • Use the qubit16 Python package to keep a durable, comparable record of every hardware run you make.
Explore other pathsStuck? Ask the tutorLook up a term