Publications Repository - Helmholtz-Zentrum Dresden-Rossendorf
FoMICS Prize for PhD Students: Presentation "Modeling ultrafast processes in laser-driven plasmas on modern compute hardware"
Conventional particle accelerators are approaching fundamental material limits which drive up machine sizes and costs. But both fundamental research and mid-energy range applications such as tumor therapy with ion beams are demanding further progress. Plasma-based accelerators driven by high-power lasers promise a fresh approach to overcome today's limits in accelerator physics. Providing a substantiation increase in accelerating gradients, this new generation of particle accelerators potentially fits on the size of a table if controlled correctly.
My PhD topic focuses on the development of open, manycore-driven particle-in-cell (PIC) simulations in order to model highly non-linear processes that happen on a time scale of atto- to pico-seconds inside this new class of particle accelerators. This includes the development of the world's fastest PIC code PIConGPU which scales up to the full size of Titan (ORNL). I am further extending the PIC algorithm to model previously unreachable multi-physics processes in high-energy density plasmas such as the interaction with XFEL laser beams at the European XFEL, enabling a new quality in predictability.
Keywords: laser plasma modeling HED atomic physics PIC GPU open source LPA ion acceleration
Platform for Advanced Scientific Computing (PASC) Conference 2017, 26.-28.06.2017, Lugano, Schweiz