GPU-accelerated 3D large-signal device simulation using the particle-in-cell code ‘Neptune’

Authors:

Abstract
We present large-signal, 3D simulations of vacuum electronic slow-wave amplifier operation using the GPU-accelerated FDTD-PIC simulation code, `Neptune'. Neptune implements a self-consistent, charge-conserving, electromagnetic, particle-in-cell algorithm optimized for parallel execution in both multi-core CPU and GPU processors. Significant performance gains achieved using a 512-core Nvidia GTX-580 graphics card are demonstrated.

This publication has 4 references indexed in Scilit: