Artificial Neural Networks have become a common solution for many real world problems. Many industrial, commercial and research applications need hardware implementation due to issues regarding stability, s peed, price and size. This paper presents the implementation of a feed forward Artificial Neu ral Network in FPGA using two embedded processors. The processors used are Xilinx hardcore PowerPCs. To verify the implementation developed, a control loop of Direct Inverse Control was simulated using a Personal Computer and a FPGA, thereby implementing a direct and an inverse model of a system, respectively. The results obtained show tha t the hardware implementation works properly and introduces no additional error. Copyri ght CONTROLO2012.

/pdf/dual-processor-neural-network-implementation-in-fpga-4u785vavj8.pdf

Dual-Processor Neural Network Implementation in FPGA

This paper proposes an efficient SIMD architecture with parallel memory for 2D cosine transforms of multiple video standards. A novel parallel memory scheme is employed to provide conflict-free parallel access in both horizontal and vertical directions with the successive or even/odd mode, as well as to eliminate data permutation and matrix transposition. Furthermore, application specific instructions are presented to accelerate the transform kernels, such as butterfly and rotate operations with scaling, rounding and clipping. The simulation results show that proposed architecture achieves significant performance improvement with low hardware cost of 3.2 K equivalent gate count for parallel memory subsystem (not including SRAMs) and 19.8 K for arithmetic units@250 MHz in 0.18 mum process.

Xing Qin

Papers

An Efficient SIMD Architecture with Parallel Memory for 2D Cosine Transforms of Video Coding