Controls and conditionals¶

Qompile has two different ways to make an operation depend on something else in the circuit - and it's easy to mix them up, because both are described using the word "control." This page explains the difference, then shows a worked example of each.

Quantum controls¶

A quantum control makes a gate depend on another qubit, while that qubit stays in superposition - nothing is measured, and no information becomes classical. The control qubit and the target stay entangled as part of one combined quantum state.

CNOT and Toffoli are the two built-in controlled gates (one and two controls, both controlling an X).
The standalone Control tile lets you build a controlled version of any gate - not just X - by connecting a control point to another gate on a different wire.

Because the control qubit isn't measured, a quantum control can act on a superposition of "control on" and "control off" at the same time - which is exactly how gates like CNOT create entanglement in the first place.

Classical conditionals¶

A classical conditional - the if tile - makes a gate depend on a classical bit, almost always one that was just set by a measurement. Because measurement has already collapsed the superposition into a definite \( 0 \) or \( 1 \) before the conditional gate runs, there's no superposition left in the decision itself - the condition is checked exactly the way an if statement in ordinary code would be.

Comparing the two¶

	Quantum control	Classical conditional
Depends on	A qubit's quantum state	An already-measured classical bit
Involves measurement?	No	Yes (upstream of the conditional)
Can act on superposition?	Yes	No - the classical bit is always definite
Built-in examples	CNOT, Toffoli	if
Custom version	Control tile, connected to any gate	if tile, wrapping any gate

Example: quantum control (no measurement)¶

H on q[0], then CNOT from q[0] to q[1]:

OPENQASM 2.0;
include "qelib1.inc";

qreg q[2];
creg c[2];

h q[0];
cx q[0], q[1];

q[0] ends up in a superposition of \( |0\rangle \) and \( |1\rangle \), so the CNOT's effect - flip q[1] when the control is 1 - happens "for both branches at once," producing the entangled Bell state \( \tfrac{1}{\sqrt{2}}(|00\rangle + |11\rangle) \). Neither qubit has a definite value yet; they're correlated, not decided.

Example: classical conditional (after measurement)¶

Measure q[0] into c[0], then apply X to q[1] only if the result was 1:

OPENQASM 2.0;
include "qelib1.inc";

qreg q[2];
creg c[2];

h q[0];
measure q[0] -> c[0];
if (c[0] == 1) x q[1];

This time, q[0] is measured before the conditional gate runs - so by the time the if is evaluated, c[0] already holds a definite 0 or 1, and q[1] either gets flipped or doesn't, with no superposition involved in that decision. This pattern - measure, then classically decide what to do next - is exactly what protocols like quantum teleportation rely on for their final correction step.