Current work:

• Ground State Energy Estimation (GSEE) using variational quantum algorithms
• Genuine Multipartite Entanglement (GME)
• Quantum algorithm development and implementation
• Quantum simulation and benchmarking across different computational platforms (CUDA-Q, QPerfect MIMIQ)
• Integration of machine learning and optimization techniques with quantum computing

Project objectives:

Improving the prediction of quantum properties of molecules with quantum algorithms, and developing and applying quantum machine learning algorithms to predict the phase equilibria of materials.

Research directions:

• Predicting materials phases and phase transitions with quantum machine learning
• Improving the precision of predicted quantum properties of large molecules (electronic structure, optical response, transport) by complementing electronic structure calculations with machine learning
• Formulating quadratic unconstrained binary optimization (QUBO) problems that represent metastable configurations of materials, and solving them on quantum annealers
• Translating classical machine learning algorithms for materials optimization into QUBO formulations solved with quantum optimization algorithms
• Wave-function-based and density-based electronic structure methods, complemented by machine learning, for predicting molecular properties

Supervisors: Prof. Guido Goldoni (UNIMORE), Prof. Rosa Di Felice (USC, USA) · Co-supervisors: Dr. Filippo Troiani (CNR), Dr. Stefano Pittalis (CNR)

Classical algorithms for ground-state energy estimation scale exponentially with system size, which makes quantum computation an attractive alternative whose results must be benchmarked on reliable data. These posters apply the Variational Quantum Eigensolver (VQE), a hybrid quantum-classical algorithm, to molecular ground states using quantum software on both CPU and GPU devices, and show the GPU acceleration of VQE.

Presented at:

• Croatia, 2025: VQE for methylene (CH₂) and benzene (C₆H₆) on CPU and GPU
• Germany, 2025: extended to DNA bases (adenine, thymine, guanine, cytosine, uracil), small organic functional groups, simple hydrocarbons and aromatic hydrocarbons up to pentacene (C₂₂H₁₄)

Abstract

Quantum Bit String Comparators (QBSC) operate on two sequences of n-qubits, enabling the determination of their relationships, such as equality, greater than, or less than. This is analogous to the way conditional statements are used in programming languages. Consequently, QBSCs play a crucial role in various algorithms that can be executed or adapted for quantum computers. The development of efficient and generalized comparators for any n-qubit length has long posed a challenge, as they have a high-cost footprint and lead to quantum delays. Comparators that are efficient are associated with inputs of fixed length. As a result, comparators without a generalized circuit cannot be employed at a higher level, though they are well-suited for problems with limited size requirements. In this paper, we introduce a generalized design for the comparison of two n-qubit logic states using just two ancillary bits. The design is examined on the basis of qubit requirements, ancillary bit usage, quantum cost, quantum delay, gate operations, and circuit complexity, and is tested comprehensively on various input lengths. The work allows for sufficient flexibility in the design of quantum algorithms, which can accelerate quantum algorithm development.

Abstract

Random walks are a prominent tool in computational mathematics for exploring extensive combinatorial structures, and the quantum walk offers a speed-up over its classical counterpart by leveraging superposition. Comparators play a pivotal role in the quantum random walk algorithm, yet existing efficient Quantum Bit String Comparators are tailored to fixed-length inputs. This thesis presents a generalized bit string comparator and a generalized discrete quantum random walk, applied to a character-movement game in which the character's motion depends on the outcome of the random walk and its target state. The designs compare two n-qubit logic states using only two ancillary bits and are evaluated for qubit requirements, ancillary bit usage, quantum cost, quantum delay, gate operations and circuit complexity across input lengths. The work also studies a superposition-based string similarity comparator requiring only log(n) qubits, and the growth of decoherence with the number of qubits in the walk.