Rolls-Royce and Classiq achieved a more than 10-fold reduction in quantum resources needed for computational fluid dynamics (CFD) tests, a critical step toward practical quantum simulations for engineering. The collaboration focused on simulating steady flow through a one-dimensional nozzle, specifically including the complex physics of transonic flow with shocks, within a hybrid classical-quantum workflow. Researchers demonstrated the CFD process could still successfully converge and produce a reliable result while utilizing an approximate quantum solver, addressing a key challenge for near-term quantum applications. “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.”
Hybrid Classical-Quantum Workflow for CFD Simulation
A tenfold reduction in required quantum resources represents a significant advancement in the pursuit of practical quantum simulations, as demonstrated by collaborative work between Classiq and Rolls-Royce. The companies’ recent investigation focused on integrating quantum computing methods into computational fluid dynamics (CFD), a computationally intensive field vital for designing complex systems across aerospace, energy, and automotive industries. Rather than seeking a perfect quantum solution, the study explored whether an approximate quantum solver could function within an established classical CFD workflow and still yield meaningful results. The classical CFD process maintained overall simulation control, while a quantum linear solver was implemented as a component within an iterative update step. Researchers found the workflow remained stable and converged, even when employing this approximate quantum solver, addressing a major concern regarding the feasibility of near-term quantum applications in engineering contexts.
This resilience suggests that fault-tolerant quantum computers may not be a prerequisite for initial benefits. The team tested a Chebyshev linear combination of unitaries (Cheb-LCU) approach, achieving the aforementioned 10x reduction in quantum resources compared to a Quantum Singular Value Transformation-based solver without sacrificing the overall convergence of the CFD process. This finding underscores the importance of evaluating quantum algorithms not in isolation, but within the broader context of a complete engineering workflow. While the initial study focused on a smaller test case, future work will concentrate on scaling the approach to tackle larger, more demanding CFD problems, potentially unlocking significant performance gains for industries reliant on complex simulations. The findings highlight a pragmatic path forward, suggesting that tolerable approximation in quantum subroutines can reduce resource demands and accelerate the adoption of quantum computing in real-world engineering applications.
Quantum computing matters to enterprises if it can fit into the workflows that engineers and researchers already use.
