Efficient Quantum Trace Estimation with Reconfigurable Real-Time Circuits

Resolution of identity constructed with quantum Hutchinson states.


Recently, quantum algorithms that leverage real-time evolution under a many-body Hamiltonian have proven to be exceptionally effective in estimating individual eigenvalues near the edge of the Hamiltonian spectrum, such as the ground state energy. By contrast, evaluating the trace of an operator requires the aggregation of eigenvalues across the entire spectrum. In this work, we introduce an efficient near-term quantum algorithm for computing the trace of a broad class of operators, including matrix functions of the target Hamiltonian. Our trace estimator is similar to the classical Girard-Hutchinson estimator in that it involves the preparation of many random states. Although the exact Girard-Hutchinson estimator is not tractably realizable on a quantum computer, we can construct random states that match the variance of the Girard-Hutchinson estimator through only real-time evolution. Importantly, our random states are all generated using the same Hamiltonians for real-time evolution, with randomness owing only to stochastic variations in the duration of the evolutions. In this sense, the circuit is reconfigurable and suitable for realization on both digital and analog platforms. For numerical illustration, we highlight important applications in the physical, chemical, and materials sciences, such as calculations of density of states and free energy.

Daan Camps
Daan Camps
Researcher in Advanced Technologies Group

My research interests include quantum algorithms, numerical linear algebra, tensor factorization methods and machine learning. I’m particularly interested in studying the interface between HPC and quantum computing.