Calibration

Calibration Protocol.jl

1using Piccolo
2using Random
3# Define system
4H_drift = PAULIS[:Z]
5H_drives = [PAULIS[:X], PAULIS[:Y]]
6sys = QuantumSystem(H_drift, H_drives, [1.0, 1.0])
7# Create trajectory
8T, N = 10.0, 100
9times = collect(range(0, T, length=N))
10pulse = ZeroOrderPulse(0.1 * randn(2, N), times)
11qtraj = UnitaryTrajectory(sys, pulse, GATES[:X])
12# Solve
13qcp = SmoothPulseProblem(qtraj, N; Q=100.0, R=1e-2)
14solve!(qcp, max_iter=100)

Calibration Protocol.py

1import pypiccolo
2import numpy as np
3# Define system
4H_drift = PAULIS['Z']
5H_drives = [PAULIS['X'], PAULIS['Y']]
6sys = QuantumSystem(H_drift, H_drives, [1.0, 1.0])
7# Create trajectory
8T, N = 10.0, 100
9times = np.linspace(0, T, N)
10pulse = ZeroOrderPulse(0.1 * np.random.randn(2, N), times)
11qtraj = UnitaryTrajectory(sys, pulse, GATES['X'])
12# Solve
13qcp = SmoothPulseProblem(qtraj, N, Q=100.0, R=1e-2)
14solve(qcp, max_iter=100)

Calibration Capabilities

Metrics

99.7%

Simulated fidelity (with decoherence)

L₁ regularization on truncated Hilbert space levels prevents leakage from model truncation artifacts

Pulses solved in rotating frame, recombined in lab frame before hardware upload

Decoherence-aware: minimum-time formulation directly reduces decoherence exposure

Model-mismatch pathway identified: iterative learning control integration planned

Result validated via photon-number-resolved spectroscopy — distinguishable per-photon peaks

A 0.9% sim-to-hardware gap. Piccolo pulses work on real hardware — not just in simulation.

Try Piccolo

Scale with Piccolismo

Calibration

Calibration Capabilities

The Solution

How are we different?

Robotics and Aerospace Algorithms

Real-time Software Design