Description:

"Optimal control of large quantum systems: assessing memory and runtime performance of GRAPE"

Yunwei Lu, PhD Student, Koch Group

Quantum optimal control is a technique used to design control pulses that can manipulate quantum systems with high accuracy. Gradient Ascent Pulse Engineering (GRAPE) is a popular method for implementing this technique, and is often combined with automatic differentiation (AD) to compute cost-function gradients. However, using AD can lead to a significant memory cost due to the storage of a large number of states and propagators. This becomes a bottleneck as the Hilbert space size of the quantum system increases. In this talk, we will revisit an alternative approach that avoids propagator storage and reduces memory requirements, by hard-coding gradients. We will benchmark runtime and memory usage, comparing this approach to AD-based implementations, with a focus on handling large Hilbert space sizes. The results demonstrate that the AD-free approach enables the application of optimal control to large quantum systems that were previously difficult to tackle.

"Cold Atoms as a Coolant for a Levitated Nanosphere"

Eduardo Alejandro, PhD Student-AMO, Geraci Group

The intermediate mesoscopic regime between classical and quantum mechanics can be explored in search of new physics using ground-state cooled silica nanospheres. In our two-chamber-trapping system, a MOT and optical tweezer prepare atoms and a nanosphere respectively for sympathetic cooling. The atoms couple to the sphere through radiation pressure forces mediated by a 1-D optical lattice. The molasses cooling of the atoms can sympathetically reduce the center-of-mass motion of the trapped sphere. Such cooled spheres can be used for precision sensing, matter-wave interferometry, and tests of quantum coherence.