Program
Full Program | Thursday | Friday | Saturday | All Sessions | Posters | Talks
LSU SQuInT Event Map
SESSION 4: Quantum simulation (Theatre)Chair: (Eleanor Rieffel (NASA Ames)) | |
| 3:30pm - 4:00pm | Nathan Lysne, Arizona Quantum simulation of complex dynamics in a quantum kicked top | Abstract. Recent advances in quantum control have enabled analog quantum simulation (AQS) as a means to study phase changes, order, and other complex many body phenomena. However, as experimental AQS grows in sophistication, new questions arise about our ability to verify the validity of a given simulation. In the absence of error correction, investigating the effects of imperfections on dynamics that is potentially chaotic and hypersensitive to errors is thus essential to understanding how much and in which ways we can trust AQS. The quantum kicked top (QKT) is an ideal model for such studies. We discuss results from recent experiments that use the d = 16 dimensional hyperfine manifold in the 6S1/2 electronic ground state of an individual Cs atom for AQS of a QKT with spin J = 15/2. As a baseline, we see close agreement between simulated and predicted dynamics in a mixed phase space over many tens of kicks. Prior work has shown the QKT dynamics reflects the separation between stable islands and sea of chaos in the classical QK, even in situations where the “fidelity” of the evolving QKT quantum state is poor. This suggests the former represents a “global” property that can be reliably simulated in the presence of errors, even when the microscopic behavior (the quantum state) cannot. We present data from experiments and numerical simulations in the presence of deliberately applied errors, showing that the frequency content of the perturbation plays a central role in the validity and robustness of AQS. |
| 4:00pm - 4:30pm | Ian Kivlichan, Harvard Bounding the costs of quantum simulation of many-body physics in real space | Abstract. Simulating many-particle dynamics, such as first-quantized quantum chemistry, with logarithmic dependence on the accuracy has proven to be a challenge. This is because the traditional approach, based on the quantum Fourier transform, introduces Hamiltonians with large max-norms. We solve this problem by using a new approach based on high-order finite difference formulae. This change makes the approach practical, and we further demonstrate that it can simulate n interacting particles using Õ(n^4) calculations of the pairwise interactions for a fixed spatial grid spacing, versus the Õ(n^5) time required by previous methods, assuming the number of particles is proportional to the number of orbitals. We also show that previous work has overlooked the fact that discretization errors can remove these exponential speedups, and address this by providing bounds on the discretization error and sufficient conditions to guarantee efficiency. |
| 4:30pm - 5:00pm | Nathan Wiebe, Microsoft Elucidating reaction mechanisms on quantum computers | Abstract. It is well known that quantum simulation promises exponential speedups for finding full configuration interaction (FCI) solutions for quantum chemistry over the best known classical algorithms. But when will this be useful? How large or a quantum computer will we need to achieve this? Here we provide estimates that show that a reasonable sized quantum computer can be used to help understand how biological nitrogen fixation works, which is a problem that requires an FCI solution. This understanding could lead to a new generation of energy efficient methods for making fertilizer that would be significant industrially. Our work considers the overheads of fault tolerance and circuit synthesis and also introduces fundamentally new circuits for simulating chemical dynamics with lower depth and introduces new methods for parallelizing phase estimation over independent quantum computers. These latter contributions help address the biggest drawback of non-variational quantum eigensolvers: their inability to be parallelized. |
SQuInT Chief Organizer
Akimasa Miyake, Assistant Professor
amiyake@unm.edu
SQuInT Co-Organizer
Mark M. Wilde, Assistant Professor LSU
mwilde@phys.lsu.edu
SQuInT Administrator
Gloria Cordova
gjcordo1@unm.edu
505 277-1850
SQuInT Event Coordinator
Karen Jones, LSU
kjones@cct.lsu.edu
SQuInT Founder
Ivan Deutsch, Regents' Professor
ideutsch@unm.edu
