XPU technology for fast and efficient FDTD simulations using modern CPUs cache memory bandwidth

EM Simulations play a key role in the design of antennas and RF components. Due to increasing complexity, higher package density or taking into account environmental aspects, the simulation effort is steadily rising, so simulation acceleration techniques become more and more important. Solutions based on graphic cards (GPU) are expensive and limited in usable memory. This paper presents the XPU technology which is a smart implementation of the Finite Difference Time Domain (FDTD) algorithm on modern CPU architectures. It can drastically reduce simulation times while (in contrast to GPU solutions) maintaining full access to the available main memory. Simulations of a 77 GHz radar sensor integrated into a car as well as a wireless vehicle charging system simulation show how the XPU technique is applied to state-of-the-art RF systems.