Quantum Phase Transitions in the Transverse-Field Ising Model: A Comparative Study of Exact, Variational, and Hardware-Based Approaches

Technical Summary: Quantum Phase Transitions in the Transverse-Field Ising Model

Problem Statement
This study addresses the challenge of characterizing quantum phase transitions (QPTs) on noisy intermediate-scale quantum (NISQ) devices. Specifically, it investigates the one-dimensional transverse-field Ising model (TFIM), a canonical system for studying quantum criticality driven by competing non-commuting interactions. While the TFIM is analytically solvable in the thermodynamic limit, finite-size systems present unique challenges: the exact ground state preserves global $Z_2$ symmetry (resulting in vanishing longitudinal magnetization), whereas variational and experimental implementations often exhibit effective symmetry breaking due to finite sampling, optimization bias, or hardware noise. The paper seeks to quantitatively assess how quantum critical behavior manifests across exact theory, ideal variational simulations, and real quantum hardware, focusing on the reliability of ground-state energies versus correlation-sensitive order parameters.

Methodology
The research employs a controlled, comparative framework using a fixed system size of $N=4$ spins and a dimensionless transverse field control parameter $h/J$. The study integrates three distinct approaches:

1. Exact Diagonalization: Serves as the symmetry-preserving benchmark. The Hamiltonian is constructed and diagonalized numerically to extract the full eigen-spectrum. To characterize magnetic order in this finite system, a correlation-based order parameter ($M_z^{exact}$) is defined, rather than simple longitudinal magnetization, to account for the preserved $Z_2$ symmetry.
2. Variational Quantum Eigensolver (VQE): A physics-inspired, depth-two ansatz is utilized, consisting of nearest-neighbor entangling operations (effective $ZZ$ interactions) followed by single-qubit $x$-axis rotations. This shallow circuit is optimized on a noiseless simulator using a gradient-free classical optimizer. The optimized parameters are then reused for both ideal simulation and hardware execution.
3. Hardware Execution: The same optimized circuits are executed on the IQM Garnet superconducting quantum processor. A resource-efficient, batched execution protocol is used to minimize queue overhead and temporal device fluctuations. Measurements are grouped by basis (Z-basis for $\langle \sigma^z_i \sigma^z_{i+1} \rangle$ and X-basis for $\langle \sigma^x_i \rangle$) to reconstruct the ground-state energy and order parameters.

Statistical uncertainties are estimated using binomial measurement statistics, and error propagation is applied to the reconstructed energy.

Key Results
The comparative analysis yields the following findings:

• Ground-State Energy: The depth-two variational ansatz accurately reproduces the exact ground-state energy across the entire parameter range, including the critical region near $h/J \approx 1$. Hardware-executed energies systematically lie above exact and ideal VQE results due to decoherence and gate imperfections but preserve the correct qualitative functional dependence on the transverse field. Energy remains a robust observable even in the presence of hardware noise.
• Magnetic Order Parameters: In contrast to energy, magnetic order parameters and correlation-sensitive observables are significantly more susceptible to noise. Ideal VQE simulations tend to overestimate magnetic order due to variational bias and symmetry breaking. Hardware results show a pronounced suppression of magnetic order across the parameter range.
• Critical Region Behavior: The quantum critical region ($h/J \in [0.8, 1.2]$) amplifies errors. On hardware, the sharp phase crossover observed in theory is broadened, and error bars increase significantly. This broadening is consistent with noise acting as an effective finite temperature, which smooths critical features and limits the system's ability to sustain long-range correlations.
• Error Metrics: Quantitative analysis (Mean Absolute Error and Root Mean Square Error) confirms that while energy errors remain bounded, order parameter errors are substantially larger, particularly for hardware results. The discrepancy between exact and hardware results is most severe for observables sensitive to long-range correlations.

Significance and Claims
The paper claims to provide a systematic, end-to-end assessment of NISQ capabilities for simulating quantum many-body physics. Its primary contributions include:

1. Differentiation of Observable Robustness: The study demonstrates that while shallow variational circuits can reliably capture ground-state energies, they struggle to faithfully reproduce symmetry-breaking order parameters and critical signatures on current hardware.
2. Noise Characterization: It highlights that quantum criticality amplifies the effects of hardware noise and finite sampling, leading to a broadening of the phase transition and the suppression of long-range correlations.
3. Benchmarking Framework: By enforcing strict consistency in system size, circuit depth, and observable definitions across exact, ideal, and hardware approaches, the work establishes a transparent methodology for isolating the effects of variational approximation from hardware noise.

The authors conclude that while NISQ devices can qualitatively capture the crossover between ferromagnetic and paramagnetic regimes, the faithful reproduction of quantum critical order remains a significant challenge. The findings underscore the necessity of careful observable selection and rigorous error analysis when interpreting hardware-based quantum simulations, suggesting that future progress requires improved hardware fidelity, more noise-resilient ansätze, and advanced error mitigation techniques.