In Vacuo and In Situ Evaluation of SIMD Codecs
In Vacuo and In Situ Evaluation of SIMD Codecs
复制标题
DOI：
10.1145/3015022.3015023
复制
发表时间：
2016-12
期刊：
影响因子：
--
通讯作者：
A. Trotman;Jimmy J. Lin
中科院分区：
文献类型：
--
作者：
A. Trotman;Jimmy J. Lin
搜索引擎索引的大小和搜索时间通过压缩编解码器密不可分。本研究使用几种相对未开发的基于SIMD的编解码器（包括QMX、TurboPackV和TurboPFor）来研究这种权衡。它使用（非SIMD）OPTPAf作为基线。介绍了QMX的四种新变体，并进行了比较。这些变体包括对空间和时间的优化。实验在TREC. gov 2集合上进行，使用主题701-850，以抓取顺序和URL顺序。结果表明，这些编解码器之间的差异很小，但QMX的参考实现表现良好。
The size of a search engine index and the time to search are inextricably related through the compression codec. This investigation examines this tradeoff using several relatively unexplored SIMD-based codecs including QMX, TurboPackV, and TurboPFor. It uses (the non-SIMD) OPTPFor as a baseline. Four new variants of QMX are introduced and also compared. Those variants include optimizations for space and for time. Experiments were conducted on the TREC .gov2 collection using topics 701-850, in crawl order and in URL order. The results suggest that there is very little difference between these codecs, but that the reference implementation of QMX performs well.