[1]Intel compilers[EB/OL].[20130516].http://software.intel.com/enus/articles/intelcompilers/.
[2]PGI compilers[EB/OL].[20130516].http://www.pgroup.com/.
[3]Gnu compiler collection[EB/OL].[20130516].http://gcc.gnu.org.
[4]Li Chunjiang, Huang Juanjuan, Xu Ying, et al. Evaluation and analysis of the effects of autovectorization in the typical compilers[J]. Computer Science, 2013, 40(4):4146.(in Chinese)
[5]Nuzman D, Henderson R. Multiplatform autovectorization[C]∥Proc of the 4th Annual International Symposium on Code Generation and Optimization,2006:281294.
[6]OpenMP 4.0 release candidate 2[EB/OL].[20130516].http://www.openmp.org/mpdocuments/OpenMP_4.0_RC2.pdf/.
[7]Huang Juanjuan, Li Chunjiang, Xu Ying. The anatomy of the cost model of autovectorization in GCC[C]∥Proc of the 17th CCF Annual Conference on Computer Engineering and Technology, 2013:259268.(in Chinese)
[8]Maddox R A, Singh G, Safranek R J. An introduction to the Intel Quick Path Interconnect[R]. CA:Intel Corporation, 2009.
[9]Manchanda N, Anand K. Nonuniform memory access(NUMA)[EB/OL].[20130516].http://cs.nyu.edu/~lerner/ spring10/ projects/NUMA.pdf.
[10]Intel64 and IA32 architectures software developer’s manual combined volumes:1, 2A, 2B, 3A and 3B[R]. CA:Intel Corporation, 2011.
附中文参考文献:
[4]李春江, 黄娟娟, 徐颖, 等. 典型编译器自动向量化效果评估与分析[J]. 计算机科学, 2013, 40(4):4146.
[7]黄娟娟,李春江,徐颖. GCC中自动向量化代价模型剖析[C]∥第17届中国计算机学会计算机工程与工艺年会论文集,2013:259268. |