首页> 外文期刊>IEEE transactions on very large scale integration (VLSI) systems >Instruction buffering to reduce power in processors for signalprocessing
【24h】

Instruction buffering to reduce power in processors for signalprocessing

机译:指令缓冲以减少处理器中用于信号处理的功率

获取原文
获取原文并翻译 | 示例

摘要

Power consumption analyzes of embedded processors indicate that ansignificant amount of power is consumed in accessing memory and in thencontrol path. Based on this, and on the runtime characteristics ofnsignal processing applications, we advocate the use of instructionnbuffering as a power-saving technique for processors for signalnprocessing and multimedia applications. Two approaches, a decodedninstruction buffer (DIB) and a decoded instruction cache, arenconsidered. Performance improvements in representative applications innspeech processing such as, the vector sum excited linear predictionn(VSELP), linear prediction coding coefficient computation (LPC), andntwo-dimensional 2-D 8×8 DCT which is used in image compression,nare provided. The reduction in power obtained is between between 25 andn30%
机译:嵌入式处理器的功耗分析表明,访问存储器和控制路径中消耗了大量的功率。基于此,并基于信号处理应用程序的运行时特性,我们提倡使用指令缓冲作为信号处理和多媒体应用处理器的节能技术。不考虑两种方法,即解码指令缓冲区(DIB)和解码指令缓存。没有提供代表性的应用程序性能改进,例如矢量和激励线性预测n(VSELP),线性预测编码系数计算(LPC)和用于图像压缩的二维2-D 8×8 DCT。所获得的功率降低介于25%至30%之间

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号