Paradata: Databases on Parallel Computers

点击翻译按钮获取中文摘要

The development of new hardware architectures encourages the development of database system technology to take advantage of this greater capability. Parallel computers hold the promise of increasing the computer throughput at lower cost than the comparable single processor system. However, these computers require that the programs that run on them be converted to exploit this parallelism. The need for ever increasing transaction rates on database systems, as well as increasing volumes of data processed by these database systems, requires research into how database systems can make effective use of parallel computer. This research will investigate how new database algorithms can achieve considerable speedup through the use of medium grained parallelism on multiple instruction stream, multiple data stream multiple instruction stream, multiple data stream (MIMD) computers.