GPGPU Powered 3D Simulations of Micro Droplets in Laser-Ion Acceleration

We present current large scale, full 3D particle-in-cell simulations and studies of laser-ion acceleration utilizing highly over-dense, mass and volume limited micro targets with PIConGPU. Powered by thousands of GPGPUs on Oak Ridge's supercomputer Titan, we show early results such as the influence of the target to laser spot size and the arising acceleration regimes thereof.

The simulations show the capability of PIConGPU, a highly scalable particle-in-cell code for many-core compute architectures that allows for in-situ, real time visualization and ultra-fast computation of large systems.