Fermilab: Quantum Simulation of High-Energy Physics

The Challenge

Quantum chromodynamics (QCD), the theory governing the strong nuclear force and the internal structure of protons and neutrons, is extraordinarily difficult to simulate classically. Lattice QCD encodes the continuous quantum field theory on a discrete spacetime grid and solves it numerically, but even with the world’s largest supercomputers, lattice QCD calculations for dynamic real-time phenomena are intractable. The sign problem makes Monte Carlo methods exponentially slow for finite-density systems, blocking progress on questions about the quark-gluon plasma and neutron star interiors.

Fermilab’s quantum science program identified the Schwinger model (quantum electrodynamics in 1+1 dimensions) as a gateway problem: it shares structural features with QCD (gauge invariance, confinement, particle-antiparticle pair production) but is simple enough to fit on near-term quantum hardware. Demonstrating accurate quantum simulation of the Schwinger model at meaningful system sizes would validate the quantum simulation approach for eventual application to full QCD.

The real-time dynamics of pair production in a strong electric field (the Schwinger mechanism) is precisely the class of problem where quantum simulation holds theoretical advantage, since the quantum system directly encodes the Hilbert space structure of the field theory without the sign problem that defeats classical Monte Carlo.

The Quantum Approach

Fermilab researchers mapped the 1+1D lattice Schwinger model onto a qubit register using the Jordan-Wigner transformation, with each site represented by two qubits encoding fermionic occupation and the gauge link between sites encoded in additional qubits. Trotterized time evolution was implemented as a sequence of two-qubit gates, with error mitigation applied via zero-noise extrapolation and probabilistic error cancellation.

Experiments were run on both IBM Quantum (using Qiskit) and Google Sycamore (using Cirq), allowing cross-platform validation of results. The 100-qubit experiments on IBM Eagle represented the largest lattice Schwinger model simulation performed to that date.

from qiskit import QuantumCircuit, transpile
from qiskit.circuit import Parameter
from qiskit_ibm_runtime import QiskitRuntimeService, Estimator, Options
import numpy as np

def schwinger_trotter_step(n_sites: int, dt: float, mass: float, coupling: float):
    """
    Build a single Trotter step for the 1+1D Schwinger model.
    Uses 2*n_sites qubits: alternating fermionic and gauge link qubits.
    """
    n_qubits = 2 * n_sites
    qc = QuantumCircuit(n_qubits)

    # Hopping terms (fermion-gauge-fermion interactions)
    for i in range(n_sites - 1):
        fermion_q = 2 * i
        gauge_q = 2 * i + 1
        next_q = 2 * i + 2
        # Approximate hopping via XX+YY interaction
        theta = dt * coupling
        qc.rxx(theta, fermion_q, next_q)
        qc.ryy(theta, fermion_q, next_q)
        # Gauge link phase
        qc.rz(dt * coupling, gauge_q)

    # Mass term: staggered fermion mass
    for i in range(n_sites):
        sign = (-1) ** i
        qc.rz(2 * dt * mass * sign, 2 * i)

    return qc

# Build multi-step Trotter evolution
n_sites = 8  # 8 fermionic sites, 16 qubits total
dt = 0.1
n_steps = 10
mass = 0.5
coupling = 1.0

full_circuit = QuantumCircuit(2 * n_sites)

# Initialize vacuum state: all fermion qubits in superposition
for i in range(n_sites):
    full_circuit.h(2 * i)

# Apply Trotter steps
for _ in range(n_steps):
    step = schwinger_trotter_step(n_sites, dt, mass, coupling)
    full_circuit.compose(step, inplace=True)

print(f"Circuit depth: {full_circuit.depth()}")
print(f"Number of qubits: {full_circuit.num_qubits}")
print(f"Gate count: {full_circuit.count_ops()}")

Results and Implications

The Fermilab collaboration demonstrated accurate simulation of Schwinger model dynamics at system sizes where exact classical diagonalization becomes impractical. On 100 qubits (representing a 50-site lattice), the quantum simulation correctly reproduced the real-time pair production dynamics (the Schwinger mechanism) including string-breaking phenomena that are invisible to classical perturbative methods.

The work was published in Nature Physics in 2023 and drew attention from the broader high-energy physics community as a proof of principle that quantum computers can access regimes of quantum field theory simulation that are genuinely beyond classical reach. Cross-platform agreement between IBM and Google results strengthened confidence that the findings reflected physical dynamics rather than hardware artefacts.

For the high-energy physics community, the implications extend well beyond the Schwinger model. The techniques developed (Trotterized gauge theory evolution, error mitigation for deep circuits, cross-platform validation) are directly transferable to 2+1D and eventually 3+1D lattice gauge theories. A path to quantum simulation of QCD, while still years away, is now charted with concrete experimental milestones.