Phase II Report¶
Parallel Performance Evaluation (Multithreaded Java Sum of Cubes)
This page includes the full standalone HTML report inside the Phase II section of the website.
Use the button below if you want to open the report as a separate full page.
Embedded Report¶
Quick Summary¶
Phase II extends the project from process-based parallelism in C to
thread-based parallelism in Java. The program computes the sum of cubes of
array elements, compares a sequential baseline to multithreaded execution,
and evaluates the required thread counts: 1, 2, 4, 6, and 8.
Key findings¶
- The best average runtime was 0.649 ms with 8 threads
- 6 threads was nearly as strong at 0.698 ms
- 2 threads performed worse than the baseline, showing clear overhead
- The major performance improvement began only at higher thread counts
What The HTML Report Adds¶
The standalone HTML report is designed to be more explanatory than the PDF. It includes:
- a plain-language walkthrough of how the Java code works
- explanation of work partitioning and thread coordination
- graphs with interpretation, not just screenshots
- benchmark evidence and discussion of why some thread counts helped more than others
- code appendix content drawn from the repository
srcfolder