Design of a reconfigurable multi-core architecture for streaming applications with a case study on performance evaluation of FIR-filters

This paper presents a reconfigurable multi-core architecture which is composed of nine nodes connected to each other in a mesh topology network. The central node hosts a general purpose processor and serves as the system controller and the surrounding nodes are equipped with coarse grain reconfigurable arrays that provide hardware flexibility and parallelism. This architecture is mainly targeted for streaming applications. As a case study, performance and hardware usage for implementations of FIR filters with different number of taps are presented. Each implementation was prototyped on an Altera FPGA and to ensure the correctness of the system, RTL simulations were carried out. The performance results were compared with the same applications implemented on a single-core processor. The timing analyses show speed-ups up to 3.6x.

[1]  Kiyoung Choi,et al.  FloRA: Coarse-grained reconfigurable architecture with floating-point operation capability , 2009, 2009 International Conference on Field-Programmable Technology.

[2]  T. Ahonen,et al.  Hierarchically Heterogeneous Network-on-Chip , 2007, EUROCON 2007 - The International Conference on "Computer as a Tool".

[3]  Jari Nurmi,et al.  Reconfigurable multi-processor architecture for streaming applications , 2012, 22nd International Conference on Field Programmable Logic and Applications (FPL).

[4]  Jari Nurmi,et al.  CREMA: A coarse-grain reconfigurable array with mapping adaptiveness , 2009, 2009 International Conference on Field Programmable Logic and Applications.

[5]  José Luis Martín,et al.  Reconfigurable Multiprocessor Systems: A Review , 2010, Int. J. Reconfigurable Comput..

[6]  Vaishali Tehre,et al.  Survey on Coarse Grained Reconfigurable Architectures , 2012 .

[7]  Jari Nurmi,et al.  Processor Design: System-On-Chip Computing for ASICs and FPGAs , 2007 .

[8]  Xin Wang,et al.  A synthesizable RTL design of asynchronous FIFO , 2004, 2004 International Symposium on System-on-Chip, 2004. Proceedings..

[9]  Cao Liang,et al.  SmartCell: An Energy Efficient Reconfigurable Architecture for Stream Processing , 2009 .