NVIDIA, Aegiq & BAE Systems test quantum boost to engineering workflows

Introduction

A joint effort between NVIDIA, Aegiq, BAE Systems and the UK National Quantum Computing Centre (NQCC) has produced a quantum‑ready architecture for large‑scale engineering simulations. By coupling NVIDIA’s DGX Spark AI infrastructure with CUDA‑Q and early‑stage quantum processors, the team demonstrated aerodynamic calculations that span billions of data points—an order of magnitude beyond typical quantum proof‑of‑concepts. The work, funded through the 2025 STFC Cross‑Cluster Proof of Concept: SparQ Quantum Computing Call, aims to show how future fault‑tolerant quantum computers could accelerate complex design tasks across aerospace, climate modeling, clean‑energy research, and more.

Hybrid Quantum‑Classical Workflow

The core of the project is a hybrid pipeline that blends classical high‑performance computing (HPC) with quantum sub‑routines. Classical nodes handle the bulk of the Navier‑Stokes discretization and data movement, while quantum processors are tasked with specific linear‑algebra kernels that are expected to benefit from quantum speed‑up (e.g., solving sparse systems or performing eigenvalue analyses). NVIDIA’s DGX Spark platform provides the orchestration layer, allowing seamless switching between CUDA‑Q kernels and traditional CUDA code. This approach lets the team scale simulations to billions of grid points without waiting for a fully fault‑tolerant quantum computer.

Engineering Impact: Aerodynamic Simulations

A key demonstration involved a two‑dimensional aerofoil flow problem. By mapping the computational domain onto a hybrid grid, the researchers achieved a full solution of the Navier‑Stokes equations at a resolution relevant to real‑world aircraft design. The quantum component accelerated the most computationally intensive linear solves, reducing overall wall‑clock time compared with a purely classical run on the same hardware. While the quantum speed‑up is modest today—reflecting the limited qubit counts and error rates of current devices—the experiment validates the integration strategy and provides a blueprint for future gains as quantum hardware matures.

Broader Applications and Outlook

The architecture is deliberately generic: any domain that relies on large‑scale PDE solvers or stochastic simulations could plug into the same workflow. Potential extensions include:

• Climate modeling – accelerating ensemble forecasts and radiative transfer calculations.
• Clean‑energy design – optimizing turbine blade geometries and battery material simulations.
• Life‑science research – speeding up molecular dynamics and protein‑folding pipelines.

Stakeholders highlighted the importance of early‑stage collaboration. BAE Systems contributed domain expertise in aerospace engineering, while NVIDIA supplied the AI‑focused hardware and software stack. Aegiq’s role centered on adapting its quantum‑ready codebase to the hybrid environment. Together, they produced a proof‑of‑concept that is both technically credible and industrially relevant.

Practical Takeaways

1. Hybrid models are the near‑term path to quantum advantage – leveraging existing HPC resources while gradually integrating quantum kernels sidesteps the wait for large, error‑corrected quantum computers.
2. Software ecosystems matter – tools like CUDA‑Q and NVIDIA’s DGX Spark simplify the development of portable quantum‑classical applications, reducing the engineering overhead for domain experts.
3. Cross‑industry partnerships accelerate adoption – combining aerospace, quantum‑software, and AI‑hardware expertise creates a feedback loop that refines both the algorithms and the hardware requirements.

Conclusion

The NVIDIA‑Aegiq‑BAE Systems collaboration demonstrates that quantum‑enhanced engineering workflows are moving from theory to practice. By successfully scaling aerodynamic simulations to billions of points with a hybrid quantum‑classical stack, the project offers a concrete roadmap for other sectors seeking to harness quantum acceleration. As fault‑tolerant quantum processors become available, the same architecture can be expanded to deliver genuine speed‑ups, potentially reshaping how complex engineering challenges are tackled across industry and research.