Israeli quantum computing startup Classiq has formed a three-way collaboration with NVIDIA and the BMW Group to tackle one of automotive engineering's harder computational problems: determining the optimal architecture for electrical and mechanical systems in electric vehicles.
The specific challenge involves selecting the best combination of components, including electric engines, batteries, and coolers, and establishing their ideal configurations relative to one another. The goal is to improve efficiency and reduce energy waste. It is the kind of high-dimensional optimisation problem that classical computing handles poorly at scale.
BMW's quantum team built an advanced quantum programme consisting of numerous sub-routines using Classiq's modelling and compilation platform. To simulate the large and complex quantum circuits this required, they drew on NVIDIA GPUs and the NVIDIA CUDA-Q platform, a hybrid classical-quantum computing environment designed for exactly this kind of workload.
"This project exemplifies the power of collaboration in the quantum computing sphere. By bringing together the best of quantum hardware, software, and automotive domain expertise, we have achieved groundbreaking advancements in a remarkably short time," said Nir Minerbi, CEO of Classiq Technologies.
The commercial implications extend beyond one manufacturer. If quantum optimisation can reliably improve mechatronic system design, it could accelerate development cycles across the electric vehicle sector and address one of the persistent concerns around EV adoption: energy efficiency under real-world operating conditions.
Significant constraints remain. Quantum hardware capable of running complex algorithms at operational scale is not yet widely available, and integrating quantum workflows into existing manufacturing processes presents logistical challenges. The cost of quantum computing resources also continues to limit accessibility. The Classiq-NVIDIA-BMW collaboration is notable precisely because it uses GPU simulation to work around hardware limitations in the near term, demonstrating a hybrid approach that offers practical progress before fault-tolerant quantum computers arrive.
The results position this collaboration as a reference point for how the automotive industry might incorporate quantum methods. The combination of domain expertise from BMW, simulation infrastructure from NVIDIA, and quantum software from Classiq produced advances that none of the three could have achieved independently in the same timeframe.