Quantum Pizza

The QSI Quantum PIzza Series is held on the 3rd Thursday of each month, from 12 - 1 pm. A diverse group of QSI Graduate Students and Postdocs will give broad overview talks on their research, or a "hot topic" paper in the quantum literature. The talks are intented to be at a very basic level -- intended for a wide audience of students, postdocs, and faculty from very different academic departments at UMBC.  Bonus -- free Pizza is served!

The speaker for the October Quantum Pizza is Anish Ghosh, a Graduate Student from the Mohammadisiahroudi Reserach Group in the Math/Stats Department.

TITLE:  Quantum Accelerated Adjoint Method For PDE-Constrained Optimization

ABSTRACT: PDE-constrained optimization problems arise in a wide range of scientific and engineering applications, including aerodynamic design, heat transfer, and inverse problems. The dominant computational cost in these problems stems from repeatedly solving large linear systems associated with discretized state and adjoint equations. Quantum linear system algorithms (QLSAs) offer exponential speedups with respect to the dimension of such systems under suitable assumptions, but their applicability to classical scientific computing is limited by the need to extract full classical solutions. In this work, we present a hybrid quantum--classical optimization framework that integrates the classical adjoint method with QLSA-based adjoint solves and swap-test-based inner product estimation. By computing reduced gradients without reconstructing the full adjoint vector, the proposed approach avoids quantum tomography and preserves the theoretical quantum advantage. By replacing the classical adjoint linear solve with a quantum solver, the complexity of our approach exhibits exponential speed-up with respect to number of state variables. Our algorithm is tested numerically using Qiskit.

Time/Location: 12:00pm, Physics Bldg. Room 401