Student: Philopatear Iweda
Mentor: Filippos Toufexis
Project Title: Streak Camera GPU Simulator
Project Description: Traditionally, picosecond and femtosecond time-resolved experiments - i.e. molecular movies - are obtained in synchrotron light sources and free-electron lasers using x-ray pulses with time duration smaller than the time resolution of the phenomenon under study. As traditional synchrotron light sources get upgraded into diffraction limited storage rings that operate with multi-picosecond pulses, doing time-resolved experiments with sub-picosecond resolution becomes impossible. We are developing a streak camera based on a high-gradient RF accelerator with theoretical femtosecond resolution. State-of-the-art streak cameras have picosecond resolution and are typically used for intensity versus time measurements. 2D movies can be inferred from the streak camera data. We need to run very efficient simulations of the system we developed to generate test cases for Machine Learning-based movie reconstruction. Therefore, we are developing a GPU simulator of the streak camera to drastically speed up the simulation time. We are looking for a student to code a fast steak camera simulator in C++ running on GPUs. It is expected that by the end of the summer, the student will generate a conference paper. Skills / Experience: (1) Strong C++ coding skills, Mathematica highly desired; (2) Experience with electromagnetic or multi-physics simulation codes such as ANSYS and CST is not necessary, but certainly desirable; (3) Excellent technical communication skills.
