Tomasz Topa

Adapting MoM With RWG Basis Functions to GPU Technology Using CUDA

In this letter, a CUDA-enabled graphics processing unit (GPU)-accelerated implementation of the method of moments (MoM) for solving three-dimensional conducting body-wire problems is presented. The solution is based on the mixed potential integral equation (MPIE) discretized using Rao-Wilton-Glisson (RWG) basis functions. The CUDA environment is employed to port a single-CPU sequential code to the parallel GPU platform, and some relevant issues are discussed. Numerical results are given for a helical antenna with a cylindrical cup reflector. The measured speedup of about eight times over the CPU implementation is demonstrated.

Absorption of EM energy by human body in the vicinity of the GSM base station antenna

A hybrid method is proposed for specific absorption rate analysis in a human body in the near field of typical GSM base-station panel antenna. The method combines the FDTD technique with an analytical description of the near field of the isolated antenna. The approximate results are found to be in excellent agreement with the results obtained by using the traditional FDTD method. The most important advantage of the proposed approach is minimization of computer memory requirements and computation time.

Using GPU accelerated version of MoM for solving scattering and radiation electromagnetic problems

In this paper, a CUDA-enabled GPU accelerated implementation of Method of Moments (MoM) for solving scattering and radiation electromagnetic problems is presented. The solution is based on the frequency-domain Electric Field Integral Equation (EFIE) formulation for arbitrary configuration of conducting bodies and wires. The CUDA environment is employed to port a single-CPU sequential code to the parallel GPU platform. Numerical results and measured speedups over CPU implementation are given for a testing structure.

Accelerating frequency-domain simulations using small shared-memory CPU/GPU cluster

Numerical approach to frequency response problems usually requires that the system governing equation is solved repeatedly at many frequencies. The computational efficiency of the overall process can be increased by departing from traditional sequential computing model in favor of utilizing the parallel processing capability commonly offered by modern hardware. In this paper, we consider a hybrid programming pattern, OpenMP + CUDA, from the perspective of a user of a rather typical low-cost multi-core CPU-based workstation that can accommodate up to four GPUs. Such the small-scale heterogeneous platforms have recently gained wide popularity in scientific computing as an inexpensive massively parallel architecture. The relevant programming model issues and performance questions are addressed. Experimental results for the example physics problem, that is, the electromagnetic scattering from perfectly electrically conducting body, show that significant performance improvement can be attained with the OpenMP + CUDA programming model.

Kernel execution strategies for GPU-accelerated version of method of moments

In this paper, a CUDA-based parallelization technique for improving performance of the Method of Moments (MoM) for solving three dimensional conducting body-wire problems is presented. The efficiency of the presented approach depends significantly on the kernel execution strategies and configuration of a GPU hardware resources specified in kernel launch. To improve the performance of the CUDA-based body-wire MoM calculations, the computation was split into several separate kernels and overlapped with the data transfer between the host and the device memories. The proposed technique is applied to analyze the radiation of the monopole antenna mounted on a PEC body and backed up with comparison with single-CPU results. A noticeable speedup (up to 10x) of the overall MoM simulation is achieved due to employing CUDA-enabled GPU.

Accelerating method of moments by using modern GPU hardware

In this paper, the GPU implementation of the method of moments (MoM) for solving electromagnetic scattering and radiation problems is presented. A single-CPU sequential MoM algorithm for analysis of arbitrary conducting 3D body-wire structures is ported using CUDA to heterogeneous CPU/GPU platform. Computational efficiency and measured speedups over a reference single-thread CPU implementation for hardware platforms with different CUDA architectures are demonstrated.

GPU-accelerated MoM-based broadband simulations using Stoer-Bulirsch algorithm

This communication introduces a CUDA-enabled GPU accelerated technique for broadband analysis of arbitrary 3D conducting body-wire radiating/scattering structure. The solution is based on the integral equation dis-cretized by the method of moments (MoM) with the use of RWG-type basis functions. Wide-band data are generated from MoM simulation employing an adaptive frequency sampling of the observed quantity, and supported by implicit rational macromodelling via Stoer-Bulirsch algorithm. Numerical results are given for a helical antenna with a cylindrical cup reflector. The GPU-CUDA offers a speedup ratio nearly 9× over the CPU implementation.