In this project, we will perform research on high-performance computing (HPC) techniques on data-intensive analysis tasks that are used at synchrotron light sources.  Specifically, we will investigate HPC problems including data access and communication patterns for iterative ptychographic and tomographic (or laminographic) reconstruction workflows on high-end systems.  Our research will also involve AI/ML usage on these workflows to enhance raw and processed data,  e.g. denoising images.  We will use large-scale experimental datasets collected at Advanced Photon Source (APS) as well as synthetic data. This project will also enable us to prepare our workflows for the upcoming Aurora system.