THRIFT - A pioneering algorithm to improve quantum simulation efficiency
Bristol, London and Washington D.C, 26 March 2025
In a breakthrough that puts us a step closer to real-world quantum applications, we have developed a novel approach to quantum simulation that significantly improves efficiency while cutting computational costs.
The method, known as THRIFT (Trotter Heuristic Resource Improved Formulas for Time-dynamics), optimises the quantum simulation according to the different energy scales found in quantum systems to more efficiently implement them into smaller, more manageable steps. This allows for larger simulations to run for longer with fewer errors, without the need to increase the size of the quantum circuits, therefore reducing the computational resources and costs involved.
In tests, as published in Nature Communications today, THRIFT improved the simulation estimates and circuit complexities of one of the most widely used benchmarks in quantum physics (1D transverse-field Ising model) by a factor of 10, allowing simulations that are 10x larger and run for 10x longer to be executed, compared with standard approaches.
Towards practical and efficient simulations
Simulating how complex systems evolve is key to designing the materials and chemicals required for challenges ranging from clean energy to advanced medicine. But these systems are constantly changing – a battery’s performance, for instance, depends on shifting quantum states, just as a drug's effectiveness relies on ever-evolving interactions between molecules.
Researchers use time-evolution simulations to model these changes. Due to the complex nature of quantum mechanics, where particles exist in multiple states at once and influence each other through entanglement, such simulations have to handle an exponential number of possibilities, often quickly overwhelming classical computers.
To tackle this, so-called Trotter formulas are used to break quantum evolution into smaller steps so they can instead be processed on today’s noisy quantum computers. Yet, even then, traditional methods not only rely on a high number of operations to maintain accuracy, the cost of which grows the longer the simulations are running, but they end up wasting time and resources by treating all interactions equally and assigning them all the same level of computational importance. This limits how long and how accurately researchers can model real-world systems, making them slow, error-prone, and impractical.
The THRIFT approach
By recognising that different interactions evolve at different speeds, THRIFT instead prioritises and channels computational power where it matters most thus reducing the number of quantum gates.
In doing so, THRIFT allows quantum computers to simulate larger systems, faster and for longer by optimising calculations, decreasing the build-up of errors, and making better use of computational resources. The longer the evolution time, the better the understanding of the properties of materials or chemicals leading to better design, and the fewer the resources, the more efficient, and ultimately practical and scalable the simulations.
Raul Santos, Lead Quantum Scientist at Phasecraft, said: “Quantum simulation is one of the most promising applications of quantum computing, but existing methods are slow, resource-intensive, and struggle to scale. With THRIFT, we've taken a significant step toward making these simulations more efficient and practical, even on today’s noisy hardware.”
Ashley Montanaro, Co-founder and CEO of Phasecraft, said: “Developing efficient quantum algorithms based on breakthrough new ideas is core to Phasecraft’s mission. Work by Raul and the team has delivered the highest-performance quantum algorithms known for simulating some prominent and well-studied physical systems. This improvement will push us closer to real-world quantum applications in materials science, chemistry, and beyond."
