IBM Quantum¶

Run QuantSDK circuits on IBM Quantum processors and the Aer simulator.

Setup¶

1. Install Dependencies¶

pip install quantsdk[ibm]

This installs qiskit, qiskit-ibm-runtime, and qiskit-aer.

2. Get an API Token¶

Create a free account at quantum.ibm.com
Go to your account settings
Copy your API token

IBM Quantum Hardware¶

Quick Start¶

import quantsdk as qs

circuit = qs.Circuit(2).h(0).cx(0, 1).measure_all()

result = qs.run(
    circuit,
    backend="ibm_brisbane",
    shots=4096,
    token="your-ibm-token"
)
print(result.counts)

Advanced Usage¶

from quantsdk.backends.ibm import IBMBackend

backend = IBMBackend(
    token="your-ibm-token",
    backend_name="ibm_brisbane",
    instance="ibm-q/open/main",
    channel="ibm_quantum_platform"
)

# Check status
info = backend.info()
print(f"Backend: {info.name}")
print(f"Qubits: {info.num_qubits}")
print(f"Status: {info.status}")

# Run
result = backend.run(circuit, shots=4096)

Available Systems¶

IBM regularly updates their fleet. Common systems include:

System	Qubits	Processor	Access
ibm_brisbane	127	Eagle r3	Open
ibm_osaka	127	Eagle r3	Open
ibm_kyoto	127	Eagle r3	Open

Check quantum.ibm.com for the current list.

Aer Simulator¶

Qiskit Aer provides high-performance local simulation with noise models.

Quick Start¶

import quantsdk as qs

circuit = qs.Circuit(2).h(0).cx(0, 1).measure_all()

# Via qs.run()
result = qs.run(circuit, backend="aer", shots=4096)

# With seed
result = qs.run(circuit, backend="aer", shots=4096, seed=42)

Direct Usage¶

from quantsdk.backends.ibm import AerBackend

aer = AerBackend(method="automatic")

info = aer.info()
print(f"Max qubits: {info.num_qubits}")

result = aer.run(circuit, shots=4096, seed=42)

Simulation Methods¶

Method	Description	Best For
`automatic`	Auto-selects best method	Default
`statevector`	Full state vector	Small circuits
`density_matrix`	Density matrix	Noise simulation
`matrix_product_state`	MPS	Deep circuits

Best Practices¶

1. Start Local, Then Go Real¶

# Development: local simulator
result = qs.run(circuit, shots=100)

# Validation: Aer with noise
result = qs.run(circuit, backend="aer", shots=4096)

# Production: real hardware
result = qs.run(circuit, backend="ibm_brisbane", shots=8192, token="...")

2. Keep Circuits Shallow¶

Hardware noise accumulates with circuit depth. Aim for:

< 100 two-qubit gates for useful results
Minimize SWAP insertions by matching your circuit to hardware topology
Use barriers to prevent unwanted optimization

3. Use Enough Shots¶

More shots = better statistics, but more QPU time:

Shots	Precision	Use Case
100	~10%	Quick test
1,024	~3%	Default
4,096	~1.5%	Good results
8,192	~1%	Production

4. Handle Queue Times¶

IBM Quantum has job queues. For time-sensitive work:

Use off-peak hours (US nighttime)
Use the Aer simulator for development
Consider IBM Premium plans for priority access

Troubleshooting¶

"No module named 'qiskit'"¶

pip install quantsdk[ibm]

"Invalid API token"¶

Verify your token at quantum.ibm.com and ensure you're using the correct channel:

backend = IBMBackend(
    token="your-token",
    channel="ibm_quantum_platform"  # or "ibm_cloud"
)

"Backend not available"¶

The system may be in maintenance. Check status at quantum.ibm.com or try a different backend.