期刊文献+

一种低功耗音频DSP的多核编程模型 被引量:1

Multi-core Programming Model for Low-power Audio DSP
下载PDF
导出
摘要 受限于低功耗音频数字DSP的多级存储层次以及有限的计算资源,编程人员在向多核扩展的过程中,面临编程困难、开销过大以及性能难以有效发挥等问题。ADMP(Audio DSP Multi-core Processing)是一种实现在APC(Audio Process-ing Core)上的单程序多数据风格的编程模型。它提供轻量级的运行时环境、显式内存分配接口,以及多种核间交互方式。4个实验用例通过ADMP均获得性能提升,实验表明ADMP以较小的开销达到了音频应用的需求。 Due to the multi-layer memory layout and constrained computing resources of low-power audio DSP, programmer suf- fers difficulties of programming, large overheads, and ineffective performances in migrating to multi-core. ADMP (audio DSP multi-core processing) is a single program multiple data style programming model on APC (audio processing core). It provides light-weight runtime, explicit memory allocation interface, and various interaction methods. Four experimental cases benefit per- forinance boost with ADMP. Experimental results show that ADMP reaches demands of audio applications with little overheads.
作者 余村 刘佩林
出处 《计算机与现代化》 2013年第4期206-209,219,共5页 Computer and Modernization
基金 国家自然科学基金资助项目(61171171)
关键词 多核 低功耗 运行时 编程模型 音频 multi-core low-power runtime programming model audio
  • 相关文献

参考文献12

  • 1Jang M, Kim K, Kim K. Tlie performance analysis of ARMNEON technology for mobile platforms [C]// Proceedings ofthe 2011 ACM Symposium on Research in Applied Compu-tation. 2011:104-106.
  • 2Khronos OpenCL Working Group. The OpenCL Specifica-tion VI. 2 [ R/OL] . http://www. khronos. org/registry/cl/, 2012-11-04.
  • 3Nvidia Corporation. CUDA Toolkit Documentation [ R/OL],http://docs. nvidia. com/cuda/index, html, 2012-10-11.
  • 4Diaz J, Munoz-Caro C, Nino A. A survey of parallel pro-gramming models and tools in the multi and many-core era[J ] , IEEE Transactions on Parallel and Distributed Sys-tems, 2012,23(8) :1369-1386.
  • 5Wei Haitao, Yu Junqing. Loading OpenMP to cell: An ef-fective compiler framework for heterogeneous multi-corechip[ C]// Proceedings of the 3rd International Workshopon OpenMP; A Practical Programming Model for the Multi-Core Era. 2008:129-133.
  • 6Lee S, Min S,Eigenmann R. OpenMP to GPGPU: A com-piler framework for automatic translation and optimization[C]// Proceedings of the 14th ACM SIGPLAN Symposiumon Principles and Practice of Parallel Programming. 2009 :101-110.
  • 7He Jiangzhou, Chen Wenguang, Chen Guangri, et al. Open-MDSP: Extending OpenMP to program multi-core DSP[C]// Proceedings of the 2011 International Conference onParallel Architectures and Compilation Techniques. 2011 :288-297.
  • 8Banakar R, Steinke S, Lee B, et al. Scratchpad memory:Design alternative for cache on-chip memory in embeddedsystems[ C ]// Proceedings of the 10th International Sym-posium on Hardware/Software Codesign. 2002:73-78.
  • 9李晋,葛敬国.Linux下互斥机制及其分析[J].计算机应用研究,2005,22(8):72-74. 被引量:12
  • 10林昊翔,秦君.Linux内核的排队自旋锁(FIFOTicketSpinlock) [ EB/OL ] . http://www. ibm. com/developer-works/cn/linu3i/l-cn-spinlock, 2008-06-05.

二级参考文献3

  • 1胡希明 毛德操.LINUX内核源代码情景分析[M].浙江大学出版社,2001..
  • 2Linus Torvalds.Linux Kernel Source Code 2.4.22[Z].
  • 3Rusty Russell.Unreliable Guide to Locking[EB/OL].http://kerneltrap.org/node/view/1776,2003-12-12.

共引文献11

同被引文献11

引证文献1

二级引证文献3

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