Insider Brief
- Classiq and Rolls-Royce published research exploring how quantum linear solvers could be integrated into computational fluid dynamics (CFD) workflows used in engineering simulations.
- The study found that a hybrid classical-quantum CFD workflow could still converge using an approximate quantum solver while significantly reducing quantum resource requirements.
- The work highlights the importance of evaluating quantum algorithms within complete engineering applications rather than as standalone components.
- Photo from Pexels by Miguel Cuenca.
PRESS RELEASE — Classiq, the leading quantum computing software company, and Rolls-Royce today published a new technical blog describing work that examines how quantum computing methods could support computational fluid dynamics (CFD), one of the most demanding areas of engineering simulation.
CFD is used across industries such as aerospace, energy, automotive and advanced manufacturing to simulate the movement of air, fluids and gases. These simulations are central to the design of aircraft, jet engines, turbines and other complex systems, but they can require significant high-performance computing resources.
The Classiq blog, “Quantum Linear Solvers for CFD: From Algorithmic Promise to Practical Performance,” looks at a practical question for future quantum computing: can a quantum linear solver be placed inside an existing CFD workflow, and can that workflow still produce useful results when the quantum component is approximate rather than perfect?
Classiq and Rolls-Royce studied a hybrid classical-quantum workflow using a CFD application made publicly available by Rolls-Royce. The application simulates steady flow through a one-dimensional nozzle, including transonic flow with shocks. In the workflow, the classical CFD process continues to manage the overall simulation, while a quantum linear solver is tested as part of an inner step that helps update the simulation.
The work found that the CFD workflow could still converge when using an approximate quantum solver. In one test, an approximate Chebyshev linear combination of unitaries, or Cheb-LCU, approach reduced quantum resource requirements by more than an order of magnitude compared with a Quantum Singular Value Transformation-based solver, while preserving convergence in the full CFD process.
The study was conducted on a smaller-scale test case, and future work will examine scaling to larger more demanding CFD problems. The findings demonstrate why testing quantum algorithms inside real application workflows is important: the practical performance of a quantum method can depend on how it behaves in the larger engineering process, not only on how it performs in isolation.
“Quantum computing matters to enterprises if it can fit into the workflows that engineers and researchers already use,” said Nir Minerbi, co-founder and CEO of Classiq. “This work is an important step in that direction. It shows how teams can move beyond evaluating algorithms on their own and begin studying how quantum methods behave inside real scientific and engineering applications.”
The blog also highlights a broader lesson for enterprise quantum teams. Future quantum applications may not require perfect quantum subroutines at every step. In some cases, useful workflows may be able to tolerate approximation if it reduces resource requirements and keeps the overall process on track.
Classiq’s role included developing and implementing the quantum portion of the hybrid CFD workflow using its high-level quantum software platform. The quantum linear solver implementation is available in Classiq’s open library, supporting repeatability and further research.
For industries that rely on simulation, this type of work can help teams prepare for future fault-tolerant quantum computers while staying grounded in real engineering needs. It also provides a practical framework for evaluating quantum methods as part of end-to-end applications rather than as standalone algorithms.
The full blog is available here: https://www.classiq.io/insights/quantum-linear-solvers-for-cfd-from-algorithmic-promise-to-practical-performance
