Curator's Take
This research tackles a fundamental challenge in quantum simulation by comparing different ways to encode molecular vibrations on quantum computers, finding that qudits (quantum systems with more than two levels) outperform traditional qubits for simulating molecules like CO2 and water. The breakthrough lies in demonstrating that qudits require fewer Hamiltonian terms, making them more resilient to noise - a critical advantage given that noise remains the biggest obstacle to practical quantum computing today. This work provides concrete evidence that moving beyond the qubit paradigm could unlock more accurate quantum simulations of molecular systems, potentially accelerating discoveries in chemistry and materials science. The findings are particularly significant because they show measurable improvements in a real application rather than just theoretical advantages, marking an important step toward quantum computers that can genuinely outperform classical methods for molecular modeling.
— Mark Eatherly
Summary
We investigate the quantum computing of the vibrational dynamics of CO$_2$ and H$_2$O by constructing the vibrational Hamiltonian in qubit and qudit form by two types of qubit encodings (binary and direct) and a qudit encoding. We simulate the time-dependent vibrational population transfer using the three different encodings, including the effect of noise and find that the qudit encoding leads to the most accurate results both for CO$_2$ and H$_2$O because of the small number of terms in the qudit Hamiltonian as long as the same values of the entangling gate error rates are adopted.