文章摘要
Xie Zhen(谢震)* **,Tan Guangming*,Liu Weifeng***,Sun Ninghui*.[J].高技术通讯(英文),2020,(3):285~298
PRF: a process-RAM-feedback performance model to reveal bottlenecks and propose optimizations
  
DOI:doi:10.3772/j.issn.1006-6748.2020.03.007
中文关键词: 
英文关键词: performance model, feedback optimization, convolution, sparse matrix-vector multiplication, sn-sweep
基金项目:
Author NameAffiliation
Xie Zhen(谢震)* ** (*Key Laboratory of Computer Architecture, Institute of Computing Technology, Chinese Academy of Sciences, Beijing 100190, P.R.China) (**University of Chinese Academy of Sciences, Beijing 100190, P.R.China) 
Tan Guangming* (*Key Laboratory of Computer Architecture, Institute of Computing Technology, Chinese Academy of Sciences, Beijing 100190, P.R.China) 
Liu Weifeng*** (***College of Information Science and Engineering, China University of Petroleum, Beijing 102249, P.R.China) 
Sun Ninghui* (*Key Laboratory of Computer Architecture, Institute of Computing Technology, Chinese Academy of Sciences, Beijing 100190, P.R.China) 
Hits: 213
Download times: 194
中文摘要:
      
英文摘要:
      Performance models provide insightful perspectives to predict performance and to propose optimization guidance. Although there has been much researches, pinpointing bottlenecks of various memory access patterns and reaching high accurate prediction of both regular and irregular programs on various hardware configurations are still not trivial. This work proposes a novel model called process-RAM-feedback (PRF) to quantify the overhead of computation and data transmission time on general-purpose multi-core processors. The PRF model predicts the cost of instruction for single-core by a directed acyclic graph (DAG) and the transmission time of memory access between each memory hierarchy through a newly designed cache simulator. By using performance modeling and feedback optimization method, this paper uses PRF model to analyze and optimize convolution, sparse matrix-vector multiplication and sn-sweep as case study for covering with typical regular kernel to irregular and data dependence. Through the PRF model, it obtains optimization guidance with various sparsity structures, algorithm designs, and instruction sets support on different data sizes.
View Full Text   View/Add Comment  Download reader
Close

分享按钮