PROJECT TITLE :
A GPU-Accelerated Parallel Shooting Algorithm for Analysis of Radio Frequency and Microwave Integrated Circuits
This paper presents a brand new parallel shooting-Newton technique based on a graphic processing unit (GPU)-accelerated periodic Arnoldi shooting solver (GAPAS) for fast periodic steady-state analysis of radio frequency/millimeter-wave integrated circuits. The new algorithm 1st explores a periodic structure of the state matrix by using a periodic Arnoldi algorithm for computing the ensuing structured Krylov subspace in the generalized minimal residual (GMRES) solver. The resulting periodic Arnoldi shooting technique is very amenable for huge parallel computing, like GPUs. Second, the periodic Arnoldi-primarily based GMRES solver within the shooting-Newton method is parallelized on the recent NVIDIA Tesla GPU platforms. We have a tendency to more explore CUDA GPUs features, such as coalesced memory access and overlapping transfers with computation to spice up the efficiency of the ensuing parallel GAPAS methodology. Experimental results from several industrial examples show that compared with the state-of-the-art implicit GMRES methodology below the identical accuracy, the new parallel shooting-Newton method can lead up to $8times$ speedup.
Did you like this research project?
To get this research project Guidelines, Training and Code... Click Here