期刊文献+

可重构算子阵列的结构和建模

Reconfigurable Operators Array:Architecture and Modeling
下载PDF
导出
摘要 提出一种由多种粗粒度、功能可配置的可重构算子组成的新型FPGA结构——可重构算子阵列结构,能完全替代细粒度的基于查找表的可编程逻辑单元,降低配置加载时间,同时电路速度具有可比性。可重构算子分为运算类、控制类、路径类和存储类,像电路指令集一样可支撑所有电路的实现。互连结构分为全局互连、局部互连和IO互连,分别承载远距离、邻近和系统外部的数据传输,互连开关采用通用开关阵列的形式。互连线段分为组线和单线两种,其中组线的位宽大于1比特,其行为一致,从而减小开关数目,提高传输速度。为了对该阵列结构进行性能分析和结构探索,还针对该结构进行建模,通过结构文件快速生成不同的结构,可作为深入研究的有效手段。 A novel FPGA fabric, reconfigurable operators (ReOps) array, which can reduce configurable time greatly and obtain a comparable performance on speed, is composed of mu.ltiple kinds of ReOps which are coarse- grain and function changeable instead of fine-grain LUT (lookup table)-based CLBs (configurable logic blocks). ReOps are classified into four kinds, arithmetic ReOps, control ReOps, path ReOps and memory ReOps. The function set of ReOps is just like a set of instruction of circuit which is soundness for circuits' implementation. A hierarchy interconnection architecture consists of global interconnection for performing long distance transfer, local interconnection for performing adjacent transfer and IO interconnection for performing system interface, is presented. The switch box is implemented by general switch array. Two kinds of segments are proposed, single line and group lines. The width of group lines is n (integer, n〉l) bits which behaved consistently and can get a great reduction on configuration bit-stream and a better speed. In order to do deep research and analysis, a model of this fabric is also developed which can obtain a different fabric quickly with different architecture file.
出处 《北京大学学报(自然科学版)》 EI CAS CSCD 北大核心 2014年第4期761-767,共7页 Acta Scientiarum Naturalium Universitatis Pekinensis
基金 深圳科技提升计划基础研究重点项目(JCYJ20120614150120560)资助
关键词 FPGA 可重构算子 结构建模 互连结构 FPGA reconfigurable operators architecture modeling interconnection architecture
  • 相关文献

参考文献3

二级参考文献16

  • 1Wang Y Y, Wang Y W. China's IC Industry Development - From the Country of Consumption to the Power Industry (in Chinese). Beijing: Science Press, Beijing, 2008. 240-241.
  • 2Coussy P, Morawiec A. High-Level Synthesis: From Algorithm to Digital Circuit. Berlin/Heidelberg: Springer Science, 2008. 16-17.
  • 3Pellerin D, Thibault S. Practical FPGA Programming in C. New Jersey: Prentice Hall, 2005. Preface: 1-2.
  • 4Gupta S, Gupta R K, Dutt N D, et al. SPARK: A Parallelizing Approach to the High-Level Synthesis of Digital Circuits. Boston: Kluwer Academic Publishers, 2004. Preface: x-xxi.
  • 5Ge L W, Chen S, Yoshimura T. Automatic implementation of arithmetic functions in high-level synthesis. In: 9th International Conference on Solid-State and Integrated Circuit Technology, Beijing, 2008. 2349-2352.
  • 6Dai P, Wang X A, Zhang X. A novel reconfigurable operator based IC design methodology for multimedia processing. In: IEEE TENCON, 2009. 1-5.
  • 7Hu Z Y, Chen K L, Wang X A. Operator design methodology and implementation for H.264 entropy encoder. In: 2nd International Conference on Information Engineering and Computer Science, Wuhan, 2010, 2. 744-745.
  • 8Richardson L E G. H.264 and MPEG-4 Video Compression. England: John Wiley & Sons Ltd, 2003. 174.
  • 9Koziri M G, Dadaliaris A N, Stamoulis G I, et al. A novel low-power motion estimation design for H.264. In: IEEE International Conference on Application Specific Systems, Architectures and Processors, Montreal, 2007. 247-252.
  • 10Tsai T H, Pan Y N. High efficiency architecture of fast block motion estimation with real-time QFHD on H.264 video coding. In: IEEE International Symposium on Multimedia, Berkeley, 2008. 124-129.

共引文献5

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部