Load balance vs. load unbalance
To reduce the communication time consumed, load balace among all the processors is important.
4 processors are used to execute the parallel Gauss elimination program on a 300 by 300 symmetric matrix. It shows, as follows, that the main part of the Gauss elimination subroutine is quite well balaced. However, an undesired initialization could cause load unbalance.
Case I: Load balance of the Gaussian elimination
The overall display window:
Case II: Load unbalance of the initialization
The overall display window: