Quantinuum’s System Model H2 quantum computer has achieved what researchers believe is the most accurate large-scale quantum simulation of Nuclear Magnetic Resonance (NMR) to date, a computationally intensive process for conventional computers, the company says. The joint team with HQS Quantum Simulations ran an end-to-end simulation reproducing key spectral features that “previous hardware demonstrations were unable to capture.” Interpreting NMR experiments becomes exponentially more expensive for classical computers as the number of interacting nuclear spins grows.
But quantum computers offer a fundamentally different approach to tackling this challenge. This advance points to new frontiers in computational chemistry and could refine analyses across fields from pharmaceutical research to materials science.
Quantinuum’s System Model H2 Enables Large-Scale NMR Simulation
Simulating a molecule notorious for challenging calculations, Quantinuum’s System Model H2 successfully modeled 1,2-di-tert-butyl-diphosphane, a benchmark compound for NMR simulations. The team’s deepest circuits reached over 1,400 two-qubit gates, simulating 70 Trotter steps that correspond to approximately 29 ms of physical evolution time. This level of complexity surpasses previous quantum hardware attempts, which failed to accurately reproduce key spectral features of the molecule. The simulation used 42 qubits (21 system qubits and 21 ancilla qubits), demonstrating a significant scale for quantum NMR.
The achievement addresses a fundamental limitation of classical computers; the computational cost of NMR simulations increases exponentially with the number of interacting nuclear spins. This poses a significant hurdle for analyzing complex molecules and materials, hindering advancements in fields like drug discovery and materials science. “Interpreting NMR experiments often requires simulations that are just as challenging as the experiments themselves,” highlighting the need for more powerful computational tools.
After applying hardware-efficient model reduction, we simulated an effective 21-spin Hamiltonian. Reproducing the characteristic double-peak structure of 1,2-di-tert-butyl-diphosphane in the simulated spectrum confirms the accuracy of the approach. “We demonstrate the most accurate large-scale digital NMR simulation performed on quantum hardware to date,” the researchers report, indicating a potential turning point for quantum chemistry. This successful end-to-end simulation, performed on System Model H2, signifies a step toward applying quantum computers to practical applications in molecular analysis and beyond.
Classical NMR Challenges and Quantum Computing Approach
This approach allowed the team to effectively represent the interactions between the molecule’s nuclear spins using fewer qubits than a direct mapping would require, mitigating the impact of qubit limitations and noise. The resulting reduction in circuit complexity was essential for achieving the fidelity needed to accurately reproduce the NMR spectrum.
Beyond reducing qubit requirements, the team’s work demonstrates a practical end-to-end workflow for quantum NMR simulation, validating algorithmic techniques important for future applications, according to Quantinuum. The approximately 29 ms of physical evolution time simulated represents an advancement in the duration of quantum simulations achievable with current technology, and is a key metric for assessing the ability of quantum computers to model dynamic chemical processes.
This achievement is particularly relevant for materials science, where understanding the behavior of molecules over time is essential for designing new materials with desired properties. As quantum hardware continues to improve, the team anticipates expanding the scope of these simulations to even more complex molecules and materials, allowing for further progress in computational chemistry and related fields.
Reproducing Key Spectral Features with 42-Qubit Computation
Simulating the behavior of 1,2-di-tert-butyl-diphosphane required over 1,400 two-qubit gates on Quantinuum’s System Model H2, a complexity that previously hindered accurate quantum reproduction of its spectral features. The team employed hardware-efficient model reduction techniques to manage the complexity of the 21-spin Hamiltonian, a technique essential for managing the computational demands of the simulation, the company says. This reduction in circuit complexity enabled the fidelity necessary to capture nuances absent in prior quantum simulations of the same molecule.
This feature had eluded capture by previous quantum hardware, demonstrating a marked improvement in simulation accuracy and scale. The successful end-to-end workflow, from initial problem setup to spectral output, establishes a practical pathway for quantum NMR. Beyond the specific molecule, this work signifies a potential shift in computational chemistry, offering a route to analyze increasingly complex materials. The exponential growth of computational cost on classical computers makes simulating molecules with many interacting nuclear spins exceptionally difficult, a limitation quantum hardware aims to overcome.
Trapped-Ion Architecture Supports Deep Quantum Circuits
Simulating 29 ms of molecular evolution, a duration critical for resolving fine spectral features in Nuclear Magnetic Resonance (NMR) analysis, was enabled by Quantinuum’s trapped-ion architecture. This achievement extends beyond prior demonstrations by completing a full computational pipeline, from Hamiltonian construction to final spectrum generation through Fourier analysis. The team, collaborating with HQS Quantum Simulations, reconstructed the free induction decay and applied error-suppression techniques during the simulation process, Quantinuum reports.
The ability to produce meaningful NMR spectra hinges on the characteristics of the underlying hardware; large Hamiltonian simulations demand both long and high-fidelity quantum circuits. Quantinuum’s system provides all-to-all qubit connectivity, high-fidelity gates and effective error-suppression techniques that preserved spectroscopic features throughout the computation. “This represents the most complete hardware demonstration of digital NMR simulation reported to date,” the researchers report, highlighting the significance of moving beyond Hamiltonian simulation alone.
The close agreement between hardware, emulator, and classical reference calculations demonstrates that deep quantum simulations of chemically meaningful systems are becoming increasingly feasible. Future utility for practicing spectroscopists will require even deeper circuits, but this work establishes that digital quantum simulation of realistic NMR experiments is no longer purely theoretical.
The experiment achieved a resolution of roughly 0.07 ppm, a promising initial step toward tackling complex molecules. Continued improvements in hardware fidelity, error correction, and quantum algorithms will be necessary to extend the simulation scale, but the current results demonstrate a path forward for interacting spin systems, among the most natural applications for quantum computers, and a shift toward practical quantum chemistry.
See today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals.
