Fewer measurements may suffice for moderate precision.

Researchers from QunaSys Inc. and Keio University are reporting a method to reduce the computational burden of quantum chemistry simulations by using a technique that effectively doubles the number of qubits used for calculations. Evaluating molecular Hamiltonians currently requires assessing O(n^4) Pauli strings, a significant hurdle for accurate estimation.

Their work numerically investigates energy estimation for molecular Hamiltonians of up to 12 qubits, demonstrating that the method requires fewer measurements than conventional approaches when targeting a precision of tens of milli-Hartree. This suggests potential benefits for applications reliant on expectation value estimation of Hamiltonians with moderate precision requirements.

Bell Sampling Reduces Measurements for Quantum Chemistry

Evaluating the absolute values of expectation values for all n-qubit Pauli strings simultaneously is now possible through a technique that uses two copies of a quantum state and coherent measurement in the Bell basis, a method explored to reduce computational demands. This approach, detailed in recent work, offers a pathway to estimate these values from a single type of quantum circuit, potentially streamlining calculations in fields like quantum chemistry and fluid dynamics.

The research focuses on diminishing the substantial measurement costs traditionally associated with accurately estimating expectation values of Hamiltonians. Conventional methods for estimating the expectation value of a Hamiltonian rely on measuring a large number of Pauli strings and subsequent classical processing, a process that scales significantly with the number of qubits involved.

To circumvent this, researchers are investigating Bell sampling, a technique that uses entangled measurements on two identical quantum states. The study numerically investigated this efficiency for molecular Hamiltonians, examining systems of up to 12 qubits. The core of this method lies in preparing two copies of the quantum state =, then performing measurements in the Bell basis. The researchers detail a concrete procedure building on previous work, to realize this idea using quantum circuits designed for these Bell-basis measurements.

While estimating the signs of these expectation values still requires separate consideration, the absolute values can be obtained with fewer quantum resources. The team’s numerical investigations focused on determining the bias and standard deviation when using this method to estimate the expectation value of a Hamiltonian expressed by Pauli strings.

This suggests a potential advantage for applications where moderate precision is sufficient, offering a trade-off between accuracy and computational cost. The researchers also explored strategies for estimating the signs of the expectation values, employing conventional sampling with a technique called QWC grouping and allocating resources based on WRS or WDS. Further analysis compared the performance of Bell sampling with conventional QWC grouping demonstrating that the absolute values are estimated by Bell sampling while the signs are estimated using conventional methods.

The results indicate that the sampling cost for sign estimation is significantly less demanding than in typical scenarios where conventional sampling is used for the entire expectation value. This division, Bell sampling for magnitudes, conventional sampling for signs, could be a key factor in reducing the overall computational burden.

The study highlights that, while conventional sampling is employed for sign estimation, it only needs to estimate the signs of the Pauli strings, making it a more efficient process. This suggests a potential pathway for optimizing the entire process, using the strengths of both Bell sampling and conventional methods. The study assesses the performance of Bell measurements on two copies of a quantum state for estimating expectation values of Hamiltonians.

The results demonstrate the potential for reducing the number of measurements required, particularly when moderate precision is acceptable. The findings suggest that this technique could be valuable for a range of applications, including quantum chemistry, physical simulations and machine learning.

Expectation Value Estimation with Two-Qubit States

The study examines the bias and standard deviation inherent in expectation value estimation, recognizing that statistical errors are a prominent obstacle to practical quantum computation. Researchers analyzed the performance of this two-copy Bell measurement method, specifically examining its ability to estimate the energy expectation value of molecular Hamiltonians. Their analysis considered the impact of the square root and max functions within the estimator, noting an unfavorable scaling of 1/ε^4 related to the max function, which will be detailed in a subsequent section.

This detailed examination of bias and standard deviation is important for understanding the method’s limitations and potential advantages in real-world applications. The researchers acknowledge that while the method may offer advantages in scaling with the number of Pauli strings, there is no guarantee it will maintain an overall advantage in energy expectation value estimation because of this 1/ε^4 scaling.

Further complicating the analysis is the potential for bias introduced during sign estimation, a necessary step to accurately calculate the Hamiltonian’s expectation value. The team’s approach involves estimating the sign using an estimator where xil = ± 1 represents the l-th measurement outcome of Pi.

The work also explores the possibility of reducing measurement overhead during sign estimation by focusing on Pauli strings with relatively large expectation values, potentially setting the expectation value of others to zero. The researchers considered the use of classical methods to approximate the signs of expectation values, offering another avenue for optimization.

“In this paper, to clarify the practical advantage of this method, we analyze the bias and the standard deviation in the expectation value estimation of molecular Hamiltonians,” the researchers state, emphasising the focus of their work on quantifying the method’s strengths and weaknesses. This research contributes to the ongoing effort to develop more efficient and accurate methods for quantum chemistry simulations. The detailed analysis of bias, standard deviation, and scaling behavior provides valuable insights for optimizing the performance of expectation value estimation techniques and for guiding future research in this critical area of quantum computation.

Bias and Standard Deviation in Energy Estimation

This advantage extends to LiH, even with variations in the active space used for calculations. However, obtaining these signs introduces an additional computational cost. The study indicates that the Bell sampling method can offer a reduction in measurement count for rough estimations, achieving accuracy up to 10^(-1) Hartree, even when accounting for the overhead of sign estimation. The efficiency of this approach hinges on carefully balancing the reduction in measurement costs with the computational demands of accurate sign estimation.

Evaluating molecular Hamiltonians necessitates assessing O(n^4) Pauli strings, a computational demand that this work addresses by using a unique approach to quantum state measurement. This simultaneous estimation is enabled by the properties of Bell states, specifically, their shared eigenstates with respect to tensor products of Pauli operators.

👉 More information
🗞 Quantum expectation value estimation by doubling the number of qubits
✍️ Hiroshi Yano, Masaya Kohda, Shoichiro Tsutsui, Ryosuke Imai, Keita Kanno, Kosuke Mitarai and Yuya O. Nakagawa
🧠 DOI: http://link.aps.org/doi/10.1103/vnfp-pk5m

Stay current

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

Avatar of Rusty Flint

Rusty Flint

Rusty is a quantum science nerd. He's been into academic science all his life, but spent his formative years doing less academic things. Now he turns his attention to write about his passion, the quantum realm. He loves all things Quantum Physics especially. Rusty likes the more esoteric side of Quantum Computing and the Quantum world. Everything from Quantum Entanglement to Quantum Physics. Rusty thinks that we are in the 1950s quantum equivalent of the classical computing world. While other quantum journalists focus on IBM's latest chip or which startup just raised $50 million, Rusty's over here writing 3,000-word deep dives on whether quantum entanglement might explain why you sometimes think about someone right before they text you. (Spoiler: it doesn't, but the exploration is fascinating)

Latest Posts by Rusty Flint: