search

ngnrsaa / qflex

Flexible Quantum Circuit Simulator (qFlex) implements an efficient tensor network, CPU-based simulator of large quantum circuits.
Apache License 2.0
97 stars 25 forks source link

Greater-than-unity amplitudes for small circuits #298

Open thisac opened 4 years ago

thisac commented 4 years ago

When constructing a small qFlex circuit and running it with the default number of cuts (which is 2), the resulting state has a greater-than-unity amplitude. This seems to be due to smaller circuits being fully separable with 2 cuts and thus, as in this case, fully disconnecting parts of the qubit-grid. A simple work-around is to manually create an ordering with no cuts.

A minimal example:

import qflexcirq
import cirq

import numpy as np

q0 = cirq.GridQubit(0, 0)
q1 = cirq.GridQubit(0, 1)
q2 = cirq.GridQubit(0, 2)
q3 = cirq.GridQubit(0, 3)

def basic_circuit():
    yield cirq.H(q0)
    yield cirq.CNOT(q0, q1)
    yield cirq.CNOT(q1, q2)
    yield cirq.CNOT(q2, q3)

circuit = cirq.Circuit()
circuit.append(basic_circuit())

#############################
## Run the QFlex simulator ##
#############################
grid = qflexcirq.QFlexGrid().create_rectangular(1, 4)
qflex_device = qflexcirq.QFlexVirtualDevice(qflex_grid=grid)
circuit = qflexcirq.QFlexCircuit(cirq_circuit=circuit, device=qflex_device)
simulator = qflexcirq.QFlexSimulator()

bitstrings = [f"{i:b}".zfill(4) for i in range(2 ** 4)]

state = simulator.compute_amplitudes(
    program=circuit,
    bitstrings=bitstrings
)

print(np.around(state, 3))

This returns/prints the unnormalized state:

[2.378+0.j 0.   +0.j 0.   +0.j 0.   +0.j 2.378+0.j 0.   +0.j 0.   +0.j 0.   +0.j 2.378+0.j 0.   +0.j 0.   +0.j 0.   +0.j 2.378+0.j 0.   +0.j 0.   +0.j 0.   +0.j]