Pennylane plugin

Damavand enters the quantum simulators community after many others. Among the most popular quantum circuit simulators we can cite PennyLane (opens new window): a python library developped by the canadian company Xanadu (opens new window) specialized in photonics hardware. PennyLane integrates gradient descent routines and many other features that makes it the most popular and efficient library to push research in the field of Quantum Machine Learning.

It was thus legitimate to integrate damavand to PennyLane through a plugin: pennylane-damavand. This page shows presents how damavand can be leveraged on supercomputers with all the functionalities of pennylane.

Gradient Descent

One of the main features of PennyLane is to integrate gradient descent schemes.

num_qubits = 10
num_layers = 5

# initialize device with damavand.qubit backend
dev = qml.device("damavand.qubit", wires=num_qubits)

@qml.qnode(dev)
def circuit():
    # build some layers
    for l in range(num_layers): 
        qml.RX(np.pi/(i+1), wires=0)
    return [qml.expval(qml.PauliZ(w)) for w in range(num_qubits)]

result = circuit()