Quantum Simulation

Ancilla and Qubit Budgets

Accounting for the extra qubits algorithms need beyond the system register, and the overhead multiplier that error correction imposes.

Where qubits go

A quantum-simulation circuit needs more qubits than the physical system it represents. The total qubit budget splits into the system register (encoding the state), algorithmic ancillas (for block-encodings, phase estimation, arithmetic), and, in the fault-tolerant setting, the enormous overhead of error correction.

Algorithmic ancillas

Kronos motion — error correction

Product formulas: essentially none, a key advantage of Trotterization.
LCU and qubitization: ancillas for PREPARE, scaling with the number of Hamiltonian terms.
Phase estimation: counting-register qubits (or one, for iterative variants).
Arithmetic and lookup: temporary registers for coherent classical computation.

The space-time trade

Ancillas often buy shallower circuits: more qubits can reduce depth and gate count, and vice versa. For example, larger PREPARE circuits with lookup tables cut T-count at the cost of extra qubits. Resource estimates explore this space-time trade-off to find the configuration that fits a given machine.

The error-correction multiplier

The dominant term is usually error correction. Each logical qubit is encoded in many physical qubits, hundreds to thousands under the surface code, depending on the physical error rate and the target logical error rate. A circuit needing a few hundred logical qubits can require millions of physical qubits once this multiplier is applied, plus the qubits devoted to magic-state distillation factories.

Managing the budget

Qubit-frugal algorithm choices (iterative phase estimation over full-register, ancilla reuse, qubit tapering by symmetry) matter most when hardware is scarce. As machines grow, the balance shifts toward minimizing runtime and T-count instead. A credible resource estimate reports both logical and physical qubit counts, because the gap between them, set by the error-correction code and hardware quality, determines whether an algorithm is runnable on a given generation of hardware. Qubit budgeting, like T-counting, is what turns algorithmic promise into an engineering forecast.