ARTÍCULO
TITULO

A Dynamic Reconfigurable Architecture for Hybrid Spiking and Convolutional FPGA-Based Neural Network Designs

Hasan Irmak    
Federico Corradi    
Paul Detterer    
Nikolaos Alachiotis and Daniel Ziener    

Resumen

This work presents a dynamically reconfigurable architecture for Neural Network (NN) accelerators implemented in Field-Programmable Gate Array (FPGA) that can be applied in a variety of application scenarios. Although the concept of Dynamic Partial Reconfiguration (DPR) is increasingly used in NN accelerators, the throughput is usually lower than pure static designs. This work presents a dynamically reconfigurable energy-efficient accelerator architecture that does not sacrifice throughput performance. The proposed accelerator comprises reconfigurable processing engines and dynamically utilizes the device resources according to model parameters. Using the proposed architecture with DPR, different NN types and architectures can be realized on the same FPGA. Moreover, the proposed architecture maximizes throughput performance with design optimizations while considering the available resources on the hardware platform. We evaluate our design with different NN architectures for two different tasks. The first task is the image classification of two distinct datasets, and this requires switching between Convolutional Neural Network (CNN) architectures having different layer structures. The second task requires switching between NN architectures, namely a CNN architecture with high accuracy and throughput and a hybrid architecture that combines convolutional layers and an optimized Spiking Neural Network (SNN) architecture. We demonstrate throughput results from quickly reprogramming only a tiny part of the FPGA hardware using DPR. Experimental results show that the implemented designs achieve a 7× faster frame rate than current FPGA accelerators while being extremely flexible and using comparable resources.

Palabras claves

 Artículos similares

       
 
Weison Lin, Yajun Zhu and Tughrul Arslan    
Edge AI accelerators are utilized to accelerate the computation in edge AI devices such as image recognition sensors on robotics, door lockers, drones, and remote sensing satellites. Instead of using a general-purpose processor (GPP) or graphic processin... ver más

 
Ernest Antolak and Andrzej Pulka    
The paper concerns research on electronics-embedded safety systems. The authors focus on the optimization of the energy consumed by multitasking real-time systems. A new flexible and reconfigurable multi-core architecture based on pipeline processing is ... ver más
Revista: Applied Sciences

 
Ye Chi, Haikun Liu, Ganwei Peng, Xiaofei Liao and Hai Jin    
Non-volatile memories (NVMs) have aroused vast interest in hybrid memory systems due to their promising features of byte-addressability, high storage density, low cost per byte, and near-zero standby energy consumption. However, since NVMs have limited w... ver más
Revista: Applied Sciences

 
Tatjana Gric and Edik Rafailov    
Tunable metamaterials belonging to the class of different reconfigurable optical devices have proved to be an excellent candidate for dynamic and efficient light control. However, due to the consistent optical response of metals, there are some limitatio... ver más
Revista: Applied Sciences

 
Shuming Yang, Changlin Xie, Yuqiang Cheng, Dianyi Song and Mengyu Cui    
The propulsion system is one of the important and vulnerable sub-systems in a strap-on launch vehicle. Among different failure modes, the thrust drop fault is the most common and remediable one. It degrades vehicle attitude tracking ability directly. To ... ver más
Revista: Aerospace