UChicago PME Quantum Science and Engineering Seminar - Kunal Sharma

This talk will discuss recent progress toward realizing quantum-centric supercomputing, where quantum processors and classical supercomputers jointly push the computational frontiers of many-body physics and chemistry. I will begin by highlighting a series of advances showing how hybrid quantum-classical workflows can already tackle correlated electron problems previously considered intractable for current quantum processors. I will then present our Sample-based Krylov Quantum Diagonalization (SKQD) algorithm — an approach to ground-state approximation that combines theoretically provable convergence with experimental scalability. Using IBM’s Heron processor, we have applied SKQD to strongly correlated impurity models involving up to 85 qubits and 6000 entangling gates, achieving accuracy comparable to density-matrix renormalization group methods. Finally, I will discuss the prospects for demonstrating quantum advantage in ground-state approximation, expectation-value estimation, and peaked-circuit experiments, and how new algorithmic ideas are needed to tolerate the noise present in current and early-FTQC processors. 

Based on the following papers: Sci. Adv. 11, eadu9991 (2025), arXiv:2501.09702, and arXiv:2508.02578