Rolls-Royce looks to quantum physics for key to high-efficiency turbines

The quest for extreme efficiency in aerospace and industrial is beginning to migrate from paper to the quantum world. Rolls-Royce has announced a collaboration with Quantinuum, Riverlane and EPCC — the UK’s National Supercomputing Centre, based at the University of Edinburgh — to explore how fault-tolerant quantum computing can transform the design of complex components such as gas turbines.

Simulating the behavior of fluids inside a turbine is one of the biggest challenges of cutting-edge engineering. As the models gain detail, the processing power required grows to the point of overloading even the most powerful supercomputers today. According to the companies, the goal of the partnership — which should extend over several years — is precisely to circumvent this bottleneck, putting quantum computers to work alongside the classical machines.

Each member enters with a piece of the puzzle. Quantinuum provides access to its quantum systems and software environment; Rolls-Royce contributes to industrial use cases and domain knowledge; Riverlane focuses on quantum error correction, a decisive step in ensuring accurate results on a large scale; and the EPCC takes care of the integration between the classical and quantum worlds, as well as the experience in supercomputing.

In practice, the group intends to test fundamental computing blocks on Quantinum’s Helios quantum computer — presented by the company as the most accurate commercial quantum processor in the world — and evaluate how these algorithms could scale in future systems, already named Sol and Apollo. The work does not start from scratch: Rolls-Royce, Riverlane and EPCC have been developing hybrid algorithms for almost five years, so far supported by classic emulators. The agreement with Quantinuum marks the passage of these experiments, previously restricted to simulation, to real quantum hardware.

The bet is based on the construction of hybrid algorithms capable of running simulations that are currently unfeasible from a computational point of view. The horizon is the so-called “teraQuOp” era, the British goal of machines capable of performing a trillion operations without errors. “Application development takes several years, and if we want to take advantage of teraQuOp devices, we need to start now,” said Leigh Lapworth, Fellow in Computational Science at Rolls-Royce.