Fault Tolerance

Quantum Error
Correction

Quantum computers are built on the most fragile objects in physics — the slightest disturbance corrupts a calculation. The trick that makes them viable is error correction: spotting and fixing mistakes without ever looking directly at the data, because looking would destroy it.

Decoherence Shor Code Steane Code Syndrome Measurement Surface Code Qubit Overhead

The Fragility Problem

A regular computer bit is sturdy — it's firmly 0 or 1, and it takes a real jolt to flip it by accident. A qubit is the opposite: it holds a delicate blend of possibilities that the faintest nudge from the outside world can ruin. Physicists call that ruin decoherence. (A qubit's superposition is disrupted by any environmental interaction.)

Decoherence creeps in two ways: a qubit can lose its energy and slump back to 0, or it can drift out of sync so the delicate timing of its superposition is scrambled. On today's leading hardware a qubit only stays coherent for millionths to thousandths of a second — far too short for serious algorithms. (T1 = energy relaxation, |1⟩ → |0⟩; T2 = dephasing, loss of phase coherence; both on the order of µs–ms for superconducting qubits.)

Worse, every operation adds its own slip. Even the best two-qubit gates miss roughly once every 100 to 1,000 tries. An algorithm like Shor's needs millions of operations — without correction, those tiny errors pile up and swamp the answer. (Two-qubit gate error rates ~0.1–1% per operation.)

🚫

You Can't Copy a Qubit

Ordinary computers fix errors by making copies — store 0 as 000, and if one bit flips, majority vote wins. Quantum mechanics flatly forbids copying an unknown quantum state, so that simple trick is off the table. (The No-Cloning Theorem.)

🔍

Looking Destroys It

You can't just inspect a qubit to see if it's broken — measuring it collapses the superposition you were trying to protect. The workaround uses helper qubits to learn what kind of error occurred without learning the data itself. (Ancilla qubits measure error syndromes, not the logical state.)

🔄

Errors Are Continuous

A classical bit either flips or it doesn't. A qubit can be knocked by any tiny amount in any direction — infinitely many possible errors. The saving insight: every possible error breaks down into a few standard types, so fixing those few covers them all. (Any error is a linear combination of Pauli operators {I, X, Y, Z}.)

Two Kinds of Mistake

Quantum errors come in two basic flavours: a bit flip (swapping the qubit's 0 and 1) and a phase flip (scrambling the relationship between them). A complete code has to catch both at once. (X: |0⟩↔|1⟩; Z: |+⟩→|−⟩.)

q₁ q₂ q₃ q₄ q₅ q₆ q₇ q₈ q₉ Block 1 Block 2 Block 3 H H H Error E Syndrome Correct H = Hadamard • = control ⊕ = CNOT target
Shor's 9-qubit code (1995) — the first quantum error-correcting code. One logical qubit is encoded across nine physical qubits in two nested layers: three inner groups of three qubits each handle bit-flip errors; the outer structure handles phase-flip errors. Any single-qubit error can be corrected. Source: Wikimedia Commons

How You Spot an Error Without Looking

The heart of the trick is to ask the right kind of question. Instead of asking "what is this qubit?" (which would destroy it), the code asks "do these qubits still agree with each other?" — a yes/no comparison that reveals an error without revealing the protected data. The pattern of yes/no answers is called the error syndrome, and it points to exactly which qubit slipped and how. (Stabilisers — multi-qubit Pauli operators that commute with the code space but anticommute with errors — are measured; the outcomes form the syndrome without collapsing the logical state.)

Take the simplest example: store the logical 0 as three identical qubits and the logical 1 likewise. Checking whether neighbouring qubits match tells you if one flipped — and which one — without ever telling you whether you're holding a 0 or a 1. Apply the matching fix and the data is restored. (Three-qubit bit-flip code: |0⟩→|000⟩, |1⟩→|111⟩; parity checks on qubits 1&2 and 2&3 locate the flip.)

Code Physical Qubits Logical Qubits Errors Corrected Notes
Shor code 9 1 Any single-qubit First QEC code; combines bit-flip + phase-flip codes
Steane code 7 1 Any single-qubit Built from a classical Hamming code (CSS code, [7,4,3])
5-qubit code 5 1 Any single-qubit Smallest possible code for single-qubit error correction
Surface code ~1,000 (per logical) 1 Many errors (high threshold ~1%) Leading candidate for fault-tolerant quantum computers
Toric / topological codes L² (lattice) 2 O(L) errors Theoretical basis for surface codes; high fault-tolerance
Planar surface code lattice diagram showing data qubits, smooth/rough boundaries, and logical operator chains
The planar surface code. Data qubits (circles) sit on lattice vertices; ancilla qubits (squares) measure X-type and Z-type stabilisers on each plaquette and vertex. Errors manifest as chains of stabiliser violations; a classical decoder identifies and corrects them. Source: LeftAsExercise

A Critical Milestone

There's a tipping point. If your hardware's error rate is below a certain threshold, then piling on more physical qubits per logical qubit drives the error rate down toward zero — redundancy wins. But if you're above the threshold, adding qubits only adds more chances for failure and makes things worse. Everything depends on which side of that line you're on. (The threshold theorem: below a code-specific physical error rate, logical error is suppressed exponentially in the number of qubits.)

Where the line sits: For the leading surface code the threshold is roughly 1% error per gate. The best processors now reach about 99.5–99.9% two-qubit gate accuracy — and as of late 2024, hardware has crossed to the good side of that line for the first time (see below). (Surface-code threshold ≈ 1% per gate.)

The catch — overhead: Protecting a single logical qubit still takes on the order of 1,000 physical qubits. Breaking RSA-2048 with Shor's algorithm needs thousands of logical qubits — translating to millions of physical ones. Today's largest processors have only about 1,000–2,000 qubits total, so the gap to a cryptographically useful machine remains enormous. (~1,000 physical per logical; ~4,000 logical → ~4 million physical for RSA-2048.)

Landmark result — Google Willow (December 2024): Google's Willow processor became the first to run a surface code below the error threshold. Encoding one logical qubit across 101 physical qubits, they showed the logical error rate dropped by about half each time they made the code bigger — and the protected qubit outlived its best individual physical qubit by 2.4×. This is the first hardware proof that adding qubits genuinely makes a quantum memory more reliable, the central promise of the threshold theorem. (Distance-7 surface code, 0.143% logical error per cycle, suppression factor Λ = 2.14, beyond breakeven; Nature 638, 920–926, 2025.)

From One Protected Qubit to Many

Willow proved a single logical qubit could be made reliable. The race through 2025 and into 2026 has been about making many of them at once — and it has played out across every major hardware type, not just Google's superconducting chips. Error correction has gone from a theory people trusted to something running on real machines everywhere. A shift in how the field measures progress came with it: rather than counting logical qubits or chasing headline "quantum advantage" demos, researchers increasingly track reliable quantum operations in complete fault-tolerant systems — the practical yardstick for a machine that actually works. ("QuOps" as the emerging success metric.)

Many Logical Qubits

Trapped-ion and neutral-atom machines reached tens of logical qubits at once: 96 on QuEra, 48 on Quantinuum's Helios, 24 on Atom Computing (with Microsoft), 12 on Infleqtion.

Quantum Zeitgeist, 2026 →

🔀

Dynamic Surface Codes

Google moved beyond Willow's fixed pattern to flexible codes that reshape how errors are caught each cycle — hexagonal, walking, and iSWAP circuits — cutting hardware cost and error sources.

Google Quantum AI, 2026 →

🏭

Systems, Not Demos

The first small fault-tolerant systems appeared — integrating many imperfect qubits into a whole that computes reliably. Success is now measured in reliable operations, not qubit counts.

Riverlane, 2026 →