Researchers Simulate Quantum Systems Using Tensor Networks

Simulating quantum systems is key for advancing quantum computing but current devices lack sufficient qubits to model complex scenarios effectively. A truncated hybrid tensor network framework, THTN, enables distributed quantum simulation across multiple smaller processors linked by conventional communication at the National University of Defence Technology. The new computational approach uses several smaller quantum processors working together instead of relying on single devices with increasingly large numbers of qubits. This framework efficiently shares calculations between these connected systems, overcoming limitations caused by current qubit counts and sidestepping issues found in existing methods like ‘circuit knitting’. By streamlining information exchange, it now allows investigation of more complex quantum models than previously achievable through simulation. The College of Computer Science and Technology devised this method for simulating quantum systems using multiple smaller processors connected by standard communication links. This avoids limitations imposed by current qubit numbers as truncated hybrid tensor networks, or THTN, efficiently distributes calculations between linked systems. A useful analogy is to consider a tensor network akin to mapping relationships within a social network: interconnected nodes represent components of the quantum system and lines show how they interact.

The team’s approach retains only the most important connections, the Schmidt modes, between different parts of the simulation, much like identifying key ingredients that define a recipe’s flavour. By streamlining information exchange, researchers can now investigate more complex models; however, questions remain about scaling this technique for even larger simulations and diverse quantum scenarios. Layered quantum system modelling enhanced by novel tensor network approach A trajectory root-mean-square error reduction of up to thirty per cent has been achieved when simulating quantum systems, surpassing previous methods utilising time-evolving block decimation (TEBD). This improvement facilitates the modelling of layered quantum structures, systems where strong interactions occur within layers but weak ones between them, which previously proved intractable due to exponential growth in computational cost related to remote gate calculations. The new truncated hybrid tensor network framework efficiently maps complex models onto a one-dimensional cut, enabling simulations with fewer qubits than before while retaining important Schmidt modes across boundaries through interface truncation. Simulations involving transverse-field Ising models on chains reaching length thirty demonstrated that accuracy was maintained even as system size increased, unlike standard TEBD which struggled at comparable parameters despite a marginally higher trajectory root-mean-square error. Truncated hybrid tensor networks (THTNs) offered an advantage when simulating systems exhibiting growing entanglement across partitions; these models tracked references matching time-evolving block decimation (TEBD) calculations whereas lower-order approximations diverged after only a few time steps. Classical simulation validated these results against TEBD using equivalent bond dimensions, with the most significant gains observed in layered models featuring strong intra-layer and weaker interlayer couplings. Benchmarks currently rely upon classical computation however, lacking demonstration of scalability to large quantum systems or accounting for decoherence effects present in physical qubits. Substantial computing power is essential for advancing materials science and drug discovery, therefore continued efforts towards simulating complex quantum dynamics remain crucial.

The team’s truncated hybrid tensor network provides an efficient method for distributing calculations, avoiding issues inherent within existing ‘circuit knitting’ protocols where computational costs escalate exponentially alongside remote gate operations; it isn’t a universal solution though. The technique performs optimally when modelling specific layered structures, prompting questions regarding its broader applicability compared with other techniques presently under development. Despite these limitations, this work represents a strong contribution toward overcoming computational bottlenecks impacting both materials science and pharmaceutical research. Representing interactions as interconnected nodes, akin to relationships in a social network, the framework distributes calculations across multiple smaller processors, circumventing the limits imposed by current qubit numbers found in single large devices through local operations linked via classical information exchange. The researchers developed a truncated hybrid tensor network that allows quantum simulations to be distributed across several smaller systems rather than requiring one very powerful device. This method improves efficiency because it avoids an exponential increase in computation time associated with existing protocols when performing remote gate operations. Classical simulation showed this approach accurately tracked results from time-evolving block decimation on chains and ladders, particularly for layered models where interactions are strong within layers but weaker between them. The authors validated their framework using equivalent bond dimensions as TEBD calculations; further work will focus on scaling the technique to larger quantum systems while accounting for decoherence effects. 👉 More information🗞 Truncated hybrid tensor networks for distributed quantum simulation✍️ Yong Liu, Guangyao Huang, Weixu Shi, Yizhi Wang, Jiandong Ouyang, Zeqian Chen and Junjie Wu🧠 ArXiv: https://arxiv.org/abs/2609.16520 More like thisPhysicsNew magnets aim to meet positron demands of 91km colliderQuantum Research NewsColumbia physicists use quantum microwaves to shield molecules from lossPhysicsFQXi finds Schrödinger’s cat in a box models quantum events in spacetimePhysicsResearchers Torque Bose-Einstein Condensate with MicrowavesStay currentSee today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals. Tags: