Thermodynamics in trapped ion quantum processors
Date issued
Authors
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
License
Abstract
As noisy intermediate-scale quantum (NISQ) processors are becoming more widely available, techniques are being developed for scalable benchmarking of such systems. Thermodynamics-based methods form a very suitable complementary toolset as they naturally scale for larger numbers of particles. The established concept of passivity has thus far not been used to set bounds on the evolution of microscopic systems initialized in thermal states. In this work, I employ two passivity-related frameworks to sense the coupling to an otherwise undetected environment, which is coined a heat leak. For the application of both frameworks, global passivity and passivity deformation, two system qubits are undergoing unitary evolution. The optional coupling to a third environmental qubit is detected as non-unitary evolution of the system qubits.
Important for the experimental realization of these thermodynamic algorithms is fast initialization of qubits in thermal (incoherent) states, which I added to the toolbox of the trapped ion platform. The employed quantum processor is based in a segmented linear ion trap, making use of high-fidelity laser-driven operations, and featuring <100 µs preparation times for multi-qubit coherent and incoherent states. As part of this work, I characterized and optimized the main entanglement-seeding operation - the light-shift gate - for robustness, resulting in a 12-hour average cycle benchmarking success rate of 99.48(5)%, enabling high-fidelity long-term measurements. Of high importance for such improvement is the addition of an active magnetic field stabilization, achieving fluctuations below 100 nT, and allowing for phase-stable measurements longer than 15 ms.
Taking advantage of the improved operation of the trapped ion quantum processor, I have realized both passivity-based algorithms. It is shown that global passivity can verify the presence of a heat leak where a test using a microscopic version of the Clausius equation fails. Passivity deformation allows for even more sensitive detection of heat leaks, and identifies a heat leak with an error margin of 5.3 standard deviations, in a scenario where both the detection based on the Clausius equation and that based on global passivity fail. My work paves the way for experimental use of passivity-based methods to characterize quantum computers in the NISQ era.