期刊文献+

支持算法组件自动替换的编程范式及编译框架 被引量:1

A programming paradigm and compiler framework for automatic replacement of algorithm components
下载PDF
导出
摘要 针对算法组件的自动替换蕴含的算法组件的兼容性判定问题,提出了一个算法组件的约束模型,从而将算法组件的兼容性判定转化为约束模型的兼容性判定。为解决转化后的兼容性判定,采用了分治思想,即只需判定原子约束强度,约束模型间的兼容性可由原子约束强度结合约束强度推导规则与放松的组件接口兼容性判定规则自动导出。为使算法自动替换更具实用性,提出了支持算法自动替换的编程范式,使得自动替换前期涉及的手工工作在编程所涉及的不同角色之间得到合理分配。设计了编译框架原型,或支持算法与编译优化选项的自动选择。实验结果显示,该系统在9个测试用例上获得的平均加速比为2.29。 Considering that automatic replacement of algorithm components implies the problem of compatibility checking of the algorithm components, a constraint model for algorithm components is proposed to transfer the compatibility checking of algorithm component into the compatibility checking of constraint models. To solve the compatibility checking, a strategy of "Divide and Conquer", is used. According to it, only the related constraint strength between atomic constraints needs to be checked. The compatibility between constraint models can be deduced from atomic constraints, by making use of the derivate rule of constraints strength and the relaxed rule of compatibility checking for component interfaces. To make it practical, a programming roles in the development of software is presented. The new paradigm that normalizes the behaviors of different paradigm helps to balance the manual work in the first stage of the replacement of the algorithm components between different roles. The proposed prototype compiler framework supports automatic selection of algorithms and compiler optimization options. Using this system, an aver- age speedup of 2.29 x is achieved on 9 benchmarks.
出处 《高技术通讯》 CAS CSCD 北大核心 2013年第11期1131-1138,共8页 Chinese High Technology Letters
基金 863计划(2012AA010902) 973计划(2011CB302504) 国家自然科学基金(60970024) 国家自然科学基金创新研究群体科学基金(60921002)资助项目
关键词 算法组件兼容性判定 编程范式 编译框架 迭代编译 compatibility checking of algorithm components, programming paradigm, compiler framework, iter-ative compilation
  • 相关文献

参考文献18

  • 1Hiral Y, Kurokawa T, Matsuo S, et al. Classification of hash functions suitable for real-life systems. IEICE Trans- actions ,2008 ,E91-A( 1 ) :64-73.
  • 2Ansel J, Chan C, Wong Y, et al. PetaBricks: a language and compiler for algorithmic choice. In: Proceedings of the 2009 ACM SIGPLAN Conference on Programming Language Design and Implementation, Dublin, Ireland, 2009.15-21.
  • 3Metzger R, Wen Z. Automatic Algorithm Recognition and Replacement. Massachusetts : MIT Press ,2000.16-18.
  • 4Thomas N, Tanase G, Tkachyshyn O, et al. A framework for adaptive algorithm selection in STAPL. In: Proceed- ings of the tenth ACM SIGPLAN symposium on Princi- ples and practice of parallel 'programming, Chicago, IL, USA, 2005. 277 -288.
  • 5Li X, Garzaran M, Padua D. A dynamically tuned sorting library. In: Proceedings of the International Symposium on Code Generation and Optimization, Palo Alto, California, 2004.111-122.
  • 6Puschel M, Moura J, Singer B, et al. Spiral : a generator for platform-adapted libraries of signal processing algo- rithms. International Journal of High Performance Compu- ting Applications ,2004,18( 1 ) :21-45.
  • 7Yu H, Zhang D, Rauchwerger L. An adaptive algorithm se- lection framework for reduction parallelization. In: Pro- ceedings of the International Conference on Parallel Ar- chitectures and Compilation Techniques, Washington, DC, USA ,2004. 278-289.
  • 8Xiong J, Johnson J, Johnson R, et al. SPL:a language and compiler for DSP algorithms. In:Proceedings of the ACM SIGPLAN Conference on Programming Language Design and Implementation, Snowbird, Utah, USA ,2001. 298-308.
  • 9Johnson T, Elgenmann R. Context-sensitive domain-inde- pendent algorithm composition and selection. In:Proceed-ings of the 2006 ACM SIGPLAN Conference on Program- ming Language Design and Implementation, Ottawa, Can- ada,2006. 151-192.
  • 10Gilat A. MATLAB : An Introduction with Applications. 4th Edition. Massachusetts :John Wiley & Sons ,2010.1-384.

同被引文献25

  • 1Esmaeilzadeh H, Blem E, AmantR S, et al. Dark silicon and the end of multicore Scaling. In: Proceedings of the 38th Annual International Symposium on Computer Archi- tecture, San Jose, USA, 2011. 365-376.
  • 2Baek W, Chihmbi T M. Green: a framework for support- ing energy-conscious programming using controlled ap- proximation. In: Proceedings of the 2010 ACM SIGPLAN Conference on Programming Language Design and Imple- mentation, Toronto, Canada, 2010. 198-209.
  • 3Sampson A, Dietl W, Fortuna E, et al. Approximate data types for safe and general low-power computation. In: Proceedings of the 32nd ACM SIGPLAN Conference on Programming Language Design and Implementation, San Jose, USA, 2011. 164-174.
  • 4Agarwal A, Rinard M, Sidiroglou S, et al. Using code perforation to improve performance, reduce energy con- sumption, and respond to failures. MIT-CSAIL-TR-2009- 042. Cambridge, MA, USA: MIT, 2009.
  • 5Ansel J, Wong Y L, Chan C, et al. Language and com- piler support for auto-tuning variable-accuracy algorithms. In: Proceedings of the 9th Annual IEEE/ACM Interna- tional Symposium on Code Generation and Optimization, Chamonix, France, 2011. 85-96.
  • 6Liu S, Pattabiraman K, Moscibroda T, et al. Flikker: Saving refresh-power in mobile devices through critical data partitioning. In : Proceedings of the 16th Internation- al Conference on Architectural Support for Programming Languages and Operating Systems, Newport Beach, USA, 2011. 213-224.
  • 7Lingamneni A, Enz C, Nagel J L, et al. Energy parsimo- nious circuit design through probabilistic pruning. In: Design, Automation & Test in Europe Conference & Exhi- bition, Grenoble, France, 2011. 1-6.
  • 8Chen T, Du Z, Sun N, et al. A Small-Footprint High- Throughput Accelerator for Ubiquitous Machine-Learning. In: Proceedings of the 19th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Salt Lake City, USA, 2014.
  • 9Esmaeilzadeh H, Sampson A, Ceze L, et al. Towards Neural Acceleration for General-Purpose Approximate Computing. In: Proceedings of the 45th Annual IEEE/ ACM International Symposium onMicroarchitecture, Van- couver, Canada, 2012. 449-460.
  • 10Du Z, Lingamneni A, Chen Y, et al. Leveraging the Er- ror Resilience of Machine-Learning Applications for De- signing Highly Energy Efficient Accelerators, In: Pro- ceedings of the 19th Asia and South Pacific Design Auto- mation Conference, Singapore, 2013. 201-206.

引证文献1

二级引证文献1

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