Quantum simulation of Fermi-Hubbard model based on transmon qudit interaction

Arian Vezvaee,Nathan Earnest-Noble,Khadijeh Najafi

2024-02-02

Abstract:The Fermi-Hubbard model, a fundamental framework for studying strongly correlated phenomena could significantly benefit from quantum simulations when exploring non-trivial settings. However, simulating this problem requires twice as many qubits as the physical sites, in addition to complicated on-chip connectivities and swap gates required to simulate the physical interactions. In this work, we introduce a novel quantum simulation approach utilizing qudits to overcome such complexities. Leveraging on the symmetries of the Fermi-Hubbard model and their intrinsic relation to Clifford algebras, we first demonstrate a Qudit Fermionic Mapping (QFM) that reduces the encoding cost associated with the qubit-based approach. We then describe the unitary evolution of the mapped Hamiltonian by interpreting the resulting Majorana operators in terms of physical single- and two-qudit gates. While the QFM can be used for any quantum hardware with four accessible energy levels, we demonstrate the specific reduction in overhead resulting from utilizing the native Controlled-SUM gate (equivalent to qubit CNOT) for a fixed-frequency ququart transmon. We further transpile the resulting two transmon-qudit gates by demonstrating a qudit operator Schmidt decomposition using the Controlled-SUM gate. Finally, we demonstrate the efficacy of our proposal by numerical simulation of local observables such as the filling factor and Green's function for various Trotter steps. The compatibility of our approach with different qudit platforms paves the path for achieving quantum advantage in simulating non-trivial quantum many-body systems.

Quantum Physics

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is to simplify the quantum simulation of the Fermi - Hubbard model by using qudit (high - dimensional qubit) in quantum computing technology. Specifically, the paper proposes a new quantum simulation method, using qudit to overcome the complexity problems encountered in the traditional qubit (two - level quantum system) - based method when simulating the Fermi - Hubbard model. This method mainly solves the following problems:

1. **Reduce the number of qubits required**: The traditional qubit method requires twice the number of qubits as the number of physical sites, and requires complex chip connections and swap gates to simulate physical interactions. However, using qudit can significantly reduce the required quantum resources because each site can be represented by a single qudit instead of two qubits.

2. **Simplify the complexity of the quantum circuit**: By taking advantage of the intrinsic symmetry of the Fermi - Hubbard model and their intrinsic relationship with Clifford algebra, the paper introduces a Qudit Fermionic Mapping (QFM) method, which can directly map fermionic degrees of freedom onto qudit, thereby simplifying the encoding cost. In addition, QFM also preserves the intrinsic connectivity of the lattice Hamiltonian, eliminating the need for any swap - like operations, which greatly reduces the number of multi - qudit gates required.

3. **Improve the efficiency of quantum simulation**: The paper shows how to verify the effectiveness and efficiency of the proposed method by numerically simulating local observables (such as filling factors and Green functions). These simulation results indicate that using qudit can achieve more efficient quantum simulation of the Fermi - Hubbard model, especially when dealing with non - trivial quantum many - body systems.

In summary, by introducing the QFM method, this paper aims to provide a more efficient and concise way to simulate the Fermi - Hubbard model, thereby providing new tools and techniques for studying strongly correlated phenomena.