Exploring the Possibility of using A GPU While Implementing Pipelining to Reduce the Processing Time in the ETL Process

Abstract

To accumulate data at one place and to make it suitable for strategic decisions we need a data warehouse system. This requires an extract, transform and load (ETL) software, which extracts data from various resources, transform it into new data formats according to required information needs, and then load it into desired data structure(s) such as a data warehouse. Such softwares take enormous time for the purpose which makes the process very slow. To deal with the problem of time taken by ETL process, parallel processing is utilised. In this paper we have proposed the use of pipelining for the parallel processing and explored the possibility of using a GPU for the process. When a computer process does not contain high parallelism it works well on CPU which contains less number of cores. Whereas the processes that contain high degree of parallelism CPU is less efficient and each independent code runs on separate core of GPU. This paper gives the basic idea of the parallel computing and also gives a simple comparison between the usage of GPU vs CPU. By comparison and analysis, we have reached a conclusion that GPU is suitable for processing large scale data parallel load where high level of parallelism is required to be run on multiple processors , however, the CPU is more suitable for processing low level parallel computing applications..