作者: Sina Meraji , John Keenleyside , Sunil Kamath , Bob Blainey
关键词: Algorithm 、 Online aggregation 、 sort 、 Aggregate (data warehouse) 、 Sargable 、 Parallel computing 、 Hash function 、 Computer science 、 Query language 、 Database 、 Query optimization
摘要: Column-store in-memory databases have received a lot of attention because their fast query processing response times on modern multi-core machines. Among different database operations, group by/aggregate is an important and potentially costly operation. Moreover, sort-based hash-based algorithms are the most common ways queries. While used in traditional Data Base Management Systems (DBMS), hash based can be applied for faster new columnar databases. Besides, Graphical Processing Units (GPU) utilized as fast, high bandwidth co-processors to improve performance The focus this article prototype operations that we created exploit GPUs. We show GPU. One parameters affect algorithm number groups hashing algorithm. get up 7.6x improvement kernel compared CPU implementation when use partitioned multi-level using GPU shared global memories.