1 / 1

Utilization of parallel b enchmarks to test p erformance of multi-core machines Student Researcher: Elakian Kanakaraj Mentor: Dr. Donald Yeung.

bluma
Download Presentation

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Utilization of parallel benchmarks to test performance of multi-core machinesStudent Researcher: ElakianKanakarajMentor: Dr. Donald Yeung Summary: I used the NU-MineBench benchmark suite, which contains twelve different data mining applications. All applications were coded in the C programming language (using an Open-MP library for parallelizing the code), but were complied with the iccintel compiler. I attempted to modify these applications, so that they could be complied using the generic gcc compiler. However, due to many incompatibility issues, I was only able to successfully migrate the APR (apriori) and HOP applications to the gcc compiler implementation. For these two applications, I tested the execution time relative to the number of threads on a machine with twelve cores (results can be seen below). Ultimately, the goal would be to use these benchmarks with machines with hundreds to thousands of cores. Results: The data collected demonstrates an expected behavior. The execution time decreases up to around twelve threads, but subsequently increases afterwards.

More Related