Skip to content
GIRIVON
Infrastructure+
Systems+
Components+
Computing Devices+
Solutions+
Compute Tools+
Knowledge+

Quantum Resource Estimator (Surface Code)

Estimate fault-tolerant resources with the standard surface-code model: choose the smallest code distance that keeps the total logical failure probability under your target, then derive physical qubits and run time.

Routing space for lattice surgery; 1 = data patches only.

Results

Code distance d
21
Logical error per qubit per round
1.00e-12
Total failure probability
2.10e-3
Physical qubits per logical qubitRotated surface code: d² data + d² − 1 measurement
881
Total physical qubits
1,32,150
Run timeLogical depth × d code cycles
21 s

These are order-of-magnitude planning estimates for research discussion, not hardware specifications.

What it calculates

  • Required code distance
  • Logical error rate per qubit per round
  • Physical qubits
  • Run time

Worked example

Using the default inputs below; change any input above to recalculate.

Inputs
Logical qubits100
Logical depth (logical time steps)1000000
Physical error rate0.001
Threshold error rate0.01
Acceptable total failure probability0.01
Layout overhead factor1.5
Magic-state factories0
Physical qubits per factory0
Code cycle time1 µs
Results
Code distance d21estimated
Logical error per qubit per round1.00e-12estimated
Total failure probability2.10e-3estimated
Physical qubits per logical qubit881calculated
Total physical qubits1,32,150estimated
Run time21 sestimated

Methodology

Below threshold, the surface code's logical error rate per qubit per round falls exponentially with distance: p_L ≈ 0.1 × (p/p_th)^((d+1)/2) (Fowler et al., 2012). Each logical time step takes about d code cycles.

The smallest odd d is chosen such that p_L × logical qubits × logical depth × d stays under the target failure probability. A rotated surface-code patch uses 2d² − 1 physical qubits; layout overhead and magic-state factories are added.

Formulas

Logical error
p_L ≈ 0.1 (p / p_th)^((d+1)/2)
Failure
P_fail ≈ p_L × k × D × d
Qubits
Q = k × (2d² − 1) × layout + factories

Assumptions

  • Uniform circuit-level noise below threshold.
  • Prefactor 0.1 and threshold 1% are common planning values; both are inputs.

Limitations

  • Real overheads depend on decoder, connectivity, T-gate count and factory design.