A researcher at Lawrence Berkeley National Laboratory has successfully simulated hadronization, the process where quarks bind together to form particles like protons and neutrons, by remotely accessing IBM quantum hardware through the Oak Ridge Leadership Computing Facility’s (OLCF) Quantum Computer User Program (QCUP). Despite physical experiments at facilities like CERN’s Large Hadron Collider providing indirect measurements, the complete steps of hadronization remain elusive, prompting the need for advanced computer simulations. This project lays the groundwork for leveraging quantum computers to perform calculations beyond the reach of even the most powerful classical supercomputers. “In principle, we know the theory that describes hadronization, but we are unable to make predictions using it because the calculations have been too difficult for a classical computer,” said Anthony Ciavarella, the Berkeley Lab research scientist who led the project; on a quantum computer, direct predictions detailing how hadronization occurs may be possible, aiding searches for new physics. IBM QCUP Enables Hadronization Simulations The ability to model the fleeting moments after high-energy particle collisions has improved thanks to a collaboration leveraging IBM quantum hardware. His findings have been published in Physical Review D. The simulation focused on the fundamental mechanism of where “strings” of gluons stretch and ultimately “snap” apart, releasing energy to create new quark-antiquark pairs. Ciavarella utilized 104 of the 156 qubits on IBM’s Heron processor, accessed via QCUP, and employed several techniques to simplify the simulation. These included focusing on heavy quarks, easier to simulate due to their limited spread, and a “scalable circuit concurrent variational quantum solver” co-developed by Ciavarella during his graduate studies. This solver prepared the quantum computer’s qubits in a stable, low-energy quantum vacuum state. “The idea is to optimize these vacuum preparation circuits on a small system size, then do it with slightly bigger systems, and then even bigger