Quantum Motion and NVIDIA are building a complete pipeline for quantum chemistry, integrating the QUICHE project with NVIDIA CUDA-Q to streamline resource estimation and simulation. This connection allows QUICHE to dispatch computations to high-performance simulators within the NVIDIA CUDA-Q platform, moving beyond theoretical compatibility to practical implementation, the company says.
The integrated workflow begins with calculations from the ORCA program package, exporting data via FCIDUMP files, and concludes with simulations accelerated by NVIDIA technology. “We want to build a single pipeline that integrates different software stacks into a single toolbox, playing to the strengths of each of the individual components,” the team explains.
QUICHE and CUDA-Q Integration for End-to-End Quantum Chemistry
This integration streamlines workflows by allowing QUICHE to interpret FCIDUMP files, the standard format for exporting active-space integrals from the ORCA program package, and translate them into instructions for the CUDA-Q environment. The use of FCIDUMP files establishes a direct data exchange method between the traditionally classical ORCA calculations and the quantum-focused QUICHE and CUDA-Q systems. This end-to-end pipeline begins with classical calculations performed by ORCA, a widely used industry standard for quantum chemistry, and extends through algorithm setup within QUICHE and NVIDIA CUDA-Q.
The resulting simulations are then executed on NVIDIA’s accelerated computing infrastructure, offering a complete pathway from initial molecular modeling to quantum simulation results. Developers report that the user experience is simplified to a concise Python notebook, delivering energies as output from ORCA input without requiring manual circuit construction or data conversion.
This streamlined approach is designed to remain consistent as quantum computing advances toward fault-tolerant systems. The team demonstrated the integration’s capabilities by simulating the spin state energetics of iron-sulfur systems, a type of chemistry with industrial relevance when scaled to larger, more complex scenarios, according to NVIDIA. “We use it to run the classical calculations and to export the active-space integrals as a standard FCIDUMP file,” explained a researcher involved in the project.
CUDA-Q Algorithms constructs the necessary qubit Hamiltonians, and the quantum phase estimation circuits generated by QUICHE are then executed as CUDA-Q kernels on NVIDIA hardware. This allows for both simulation and resource estimation, providing a pathway to tackle chemical problems exceeding the limits of classical computation. The pipeline’s design prioritizes scalability, with the same QUICHE specification used for both simulation and resource estimation, ensuring a consistent workflow as quantum computers evolve.
The developers anticipate that this integration will facilitate the exploration of increasingly complex chemical systems and provide accurate resource estimates for simulations that are currently intractable for classical computers. A full demonstration of the integrated workflow is available for review, alongside blog posts detailing ongoing work in fault-tolerant compilation and quantum chemistry state preparation.
The ability to estimate resources alongside simulation results is a key benefit, allowing researchers to assess the feasibility of tackling increasingly complex chemical problems. The team selected the spin state energetics of Fe(SH)4− as a representative example, noting its potential relevance to industrial applications when scaled up.




See today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals.
