Evolution of Memory Mapped Files on Dual Core Processors using Large Data Mining Data Sets

摘要

In the recent years major CPU designers have shifted from ramping up clock speeds to add on-chip multicore processors. A study is carried out with data mining (DM) algorithms, to explore the potential of multi core hardware architecture with OpenMP. The concept of memory mapped files is widely supported by most of the modern operating systems. Performance of memory mapped files on multicore processor is also studied. In our experiments popular clustering algorithms such as k-means [21], max-min [22] are used. Experiments are carried out with serial versions and parallel versions. Experiment results with both simulated and real data demonstrates the scalability of our implementation and effective utilization of parallel hardware, which benefits DM problems, involves large data sets.