The minimum number of extra qubits needed for reliable quantum calculations has reduced with optimised quantum circuit designs that cut failure rates by a factor of three to eighteen times. Scalable tensor network techniques extending previous capabilities used up to fifty ancilla qubits alongside the release of a database containing these resource-efficient state preparation circuits. New methods improving tapered Quantum Phase Estimation (tQPE) developed increasing accuracy by reducing the number of ‘ancilla’ qubits required for support.
The team created optimised circuit designs and released a freely available database for other scientists. Testing on existing quantum computers demonstrated reductions in failure rates, making tQPE more practical for complex problems. New methods improve tapered Quantum Phase Estimation (tQPE), enabling more accurate estimations with fewer supporting ‘ancilla’ qubits. These extra ancillas assist calculations without forming part of the final result itself.
But how easily can these new techniques be integrated into larger quantum algorithms and will this sharply accelerate progress in areas such as materials science or drug discovery.
Scalable Tensor Networks Enable Fifty Qubit State Preparation with Reduced Error Rates
Failure probabilities dropped between three and eighteen times on Quantinuum’s H2-2 computer when using tensor network methods; this improvement occurred for values of *l* ranging from one to ten compared with standard approaches. Scalable tensor networks overcame this limitation, extending capabilities to fifty ancilla qubits while also releasing a database containing resource-efficient designs. Experiments then utilised these optimised state preparations on the H2-2 computer, again demonstrating failure probability reductions between three and eighteen times for *l* values from one to ten when contrasted with standard methods.
The team further compiled circuits down to single-qubit O gates, revealing that their approach requires fewer T-gates than existing techniques, scaling logarithmically instead of linearly with increasing precision qubits at fixed error rates. However, despite achieving an impressive ε
Reduced errors advance foundational techniques for material simulation and pharmaceutical design
Initial tests demonstrably cut failure rates using these new methods on Quantinuum’s hardware; however, a significant practical hurdle remains, the tensor network approach currently excels at preparing states for estimating only *one* property of a system. Many real world applications require simultaneous estimation of multiple properties, which this work does not yet address in terms of efficiently extending state preparation circuits beyond single parameter estimations. Despite this limitation, the advances matter as they have improved performance by reducing error rates within tapered quantum phase estimation (tQPE), a technique fundamental to many potential applications including materials science and drug discovery.
This research developed scalable tensor network methods for creating optimized state preparation circuits used in tapered quantum phase estimation (tQPE). This matters because it reduces failure probabilities, by as much as eighteen times in initial tests on Quantinuum’s H2-2 computer, while also requiring fewer computational gates than previous approaches. The team created resource-efficient circuits applicable to up to fifty ancilla qubits and released an accompanying database for others to utilize. These improvements allow tQPE, a technique important for fields like materials science and drug discovery, to function more efficiently with reduced qubit requirements.
👉 More information
🗞 Scalable Compilation of Resource-Efficient Tapers for High-Confidence Quantum Phase Estimation
✍️ David Raveh, Andrew Sornborger, Yiğit Subaşı and Lukasz Cincio (Affiliation: Los Alamos National Laboratory)
🧠 ArXiv: https://arxiv.org/abs/2611.15076




See today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals.
