Breaking through quantum bottlenecks: how classical shortcuts enable more complex materials simulations 

We demonstrated a novel approach to quantum simulation through the Hartree National Centre for Digital Innovation (HNCDI) in collaboration with IBM and the National Physical Laboratory, pushing the boundaries of what’s possible on today’s quantum hardware. 

When preparation becomes the bottleneck 

Understanding how materials respond to sudden environmental changes, such as a magnetic field shift or rapid cooling, is fundamental to designing things such as next-generation electronics and energy storage systems. These sudden changes are called “quenches” because they happen so fast that the material can’t equilibrate gradually. We expect quantum computers to excel at simulating these materials by directly representing how particles behave according to quantum physics, supporting their development. However, there’s a critical roadblock. Preparing the initial quantum state often consumes so much of the quantum computer’s limited resources that little capacity remains for the actual simulation. 

Think of it like spending most of your fuel just getting to the starting line of a race. For many materials science simulations, the initial entangled quantum state, or “starting line” requires such deep quantum circuits that only the simplest, shortest simulations can follow. This has severely limited the complexity of quantum materials that researchers can study on pre-fault-tolerant quantum hardware. 

Hybrid solutions offering the best of both worlds 

Our team, working with IBM and partners from the National Physical Laboratory have demonstrated a powerful solution, using classical supercomputing to dramatically reduce the quantum resources needed for state preparation. 

The approach works because many of these initial quantum states can be efficiently represented on classical computers using tensor networks. Tensor networks are a way to approximate quantum systems which otherwise require huge amounts of data by connecting many small blocks of information instead of storing one massive object. By using classical computation to discover “shortcut” quantum circuits, we preserve the quantum hardware’s capacity for where it matters: simulating the most complex dynamics. 

We developed two complementary approaches. The first enhances an existing algorithm called AQC-Tensor with a smarter starting point. The second, ADAPT-AQC, is a new method that intelligently builds circuits step-by-step, selecting each quantum operation based on what will most improve the result. Both methods use classical optimisation to find much shorter quantum circuits than traditional approaches 

Validation on quantum hardware 

To validate this approach at scale, we simulated the dynamics of a 50-site quantum magnet (a chain of 50 interconnected magnetic particles, each represented by one qubit) on IBM’s quantum hardware, specifically studying how an antiferromagnetic material responds to a sudden change in magnetic field (a “quantum quench”). 

The results demonstrate significant practical improvements: 

• 81-89% shallower circuits compared to standard methods: this means far less time for errors to accumulate, making the simulation much more accurate on noisy quantum hardware 

• Successfully handled high complexity: we executed circuits with 59 layers of quantum operations and 1,251 gates total, pushing the boundaries of what current classical computers can reliably simulate 

• High-fidelity results: our simulations matched the predictions from classical calculations within statistical uncertainty, confirming the quantum computer produced trustworthy answers 

Previous quantum hardware studies were limited to starting from simple, unentangled states. Our work demonstrates dynamics from a realistic, complex starting state, one that even powerful classical supercomputers struggle to simulate accurately. 

Why this matters 

This work joins recent quantum-centric supercomputing advances that demonstrate a practical way forward for near-term quantum computing: integrating classical and quantum resources synergistically rather than treating them as competing approaches. The ability to simulate realistic quantum quenches is directly relevant to understanding phase transitions in functional materials such as superconductors, magnetic storage materials and battery components, relaxation dynamics in quantum systems, and quantum effects in chemical reactions. 

As quantum hardware improves, this hybrid approach could tackle materials with complexity beyond classical capabilities, potentially enabling simulations of more complex materials that are important for future technologies. 

This research opens new possibilities for quantum simulation of realistic materials. Read the full paper here or explore our wider work with quantum computing.