Benchmarking the Performance of NISQ-Era Quantum Processors in Simulating Molecular Structures for Drug Discovery

Abstract

The application of quantum computing to molecular simulation promises to revolutionize computational chemistry and drug discovery by providing exact solutions to the electronic Schrödinger equation, a task that is intractable for classical computers for large molecules. However, the current era of Noisy Intermediate-Scale Quantum (NISQ) hardware is characterized by processors with a limited number of qubits, high error rates, and short coherence times, posing significant challenges for practical quantum advantage. This study presents a comprehensive                           benchmarking framework to                           evaluate the performance of leading NISQ-era superconducting quantum processors from IBM and Rigetti in simulating fundamental molecular structures relevant to drug discovery. We focus on the variational quantum eigensolver (VQE) algorithm to compute the ground-state energy of a series of prototypical molecules—hydrogen (H₂), lithium hydride (LiH), and a prototypical drug fragment, formic acid (HCOOH)—each increasing in complexity and qubit requirement. Our benchmarking suite assesses not only the accuracy of the computed energy relative to the Full Configuration Interaction (FCI) baseline but also critical hardware-performance metrics including algorithm success probability, convergence stability, and required quantum volume. The results demonstrate that while small molecules like H₂ (4 qubits) can be simulated with reasonable accuracy (error < 5 kcal/mol) on current hardware, the simulation of LiH (12 qubits) and HCOOH (16+ qubits) is severely hampered by noise, with 
errors exceeding chemical accuracy (1.6 kcal/mol) by an order of magnitude. Furthermore, we quantify the exponential growth in the number of required two-qubit gates and the corresponding fidelity loss, identifying this as the primary bottleneck. Through randomized compilation and noise-aware mapping, we show a 40% reduction in error for the H₂ simulation but minimal improvement for larger molecules, underscoring the fundamental limitations of NISQ devices. This work provides a rigorous, quantitative assessment of the current capabilities and limitations of quantum hardware for quantum chemistry, establishing a clear benchmark for progress and suggesting that fault-tolerant quantum computing, rather than NISQ-era devices, will likely be necessary for impactful quantum-driven drug discovery.