Physic Labs

Frontier physics

Quantum error correction and decoherence

Study the three-qubit repetition code: syndrome measurement locates a bit-flip error without disturbing the encoded state, and verify the decode-failure rate pfail=3e2−2e3p_{fail}=3e^2-2e^3 drops below pfail=ep_{fail}=e when the error probability e is small. Vary the noise level to see decoherence and the code's break-even point.

Research

Equipment

  • Three encoded qubits with a virtual syndrome table
  • Noise slider (bit-error probability e)
  • Parameter and physical-parameter sliders of the model
  • Readout of $p_{fail}=3e^2-2e^3$ and decoding fidelity

Procedure

  1. See errors detected by the syndrome

    Keep the noise low: on the canvas a random bit-flip occasionally flips one of three qubits, and the syndrome (disagreement among qubit pairs) reveals the error's location without reading the encoded value — so the quantum superposition is preserved.

  2. Measure the decode-failure rate

    Raise the noise e and read the result: the code corrects one error, but two simultaneous errors defeat the majority vote, giving pfail=3e2−2e3p_{fail}=3e^2-2e^3. At e = 1% we get pfail≈0,03%p_{fail}\approx0{,}03\% — about 30 times better than a bare qubit.

  3. Find the code's break-even point

    Keep raising e into the tens of percent: the fidelity curve shows that above e > 1/2 the code corrupts more than it corrects. Estimate the crossing pfail=ep_{fail}=e from the plot (e = 1/2) and discuss why practical decoders need physical error rates below ~10% on larger surface codes.

Simulation

Experiment history

The fragility of quantum states — decoherence — was analyzed by Erich Joos and H. Dieter Zeh in the early 1970s, and many thought it forbade quantum computing. Peter Shor proved otherwise in 1995: nine qubits could encode one logical qubit protected against both bit and phase errors, founding quantum error correction. Andrew Steane gave the 7-qubit CSS code in 1996, and Kitaev, Bravyi, and Freedman–Meyer built the topological (surface) codes of the late 1990s — now the dominant architecture for fault-tolerant quantum computers at Google and IBM. In 2023–2024 Google Quantum AI showed that larger surface codes lower the logical error rate — the first milestone of 'below-threshold' error correction.

Related physicists

Related library topics