Finite difference time domain (FDTD) method is a very poplar way of numerically solving partial differential equations. FDTD has\na low operational intensity so that the performances in CPUs and GPUs are often restricted by the memory bandwidth. Recently,\ndeeply pipelined FPGA accelerators have shown a lot of success by exploiting streaming data flows in FDTD computation. In spite\nof this success, many FPGA accelerators are not suitable for real-world applications that contain complex boundary conditions.\nBoundary conditions break the regularity of the data flow, so that the performances are significantly reduced. This paper proposes\nan FPGA accelerator that computes commonly used absorbing and periodic boundary conditions in many 3D FDTD applications.\nAccelerator is designed using a ââ?¬Å?C-likeââ?¬Â programming language called OpenCL (open computing language). As a result, the\nproposed accelerator can be customized easily by changing the software code. According to the experimental results, we achieved\nover 3.3 times and 1.5 times higher processing speed compared to the CPUs and GPUs, respectively. Moreover, the proposed\naccelerator ismore than 14 times faster compared to the recently proposed FPGA accelerators that are capable of handling complex\nboundary conditions.
Loading....