[ztdep]

Dear friends:
I am doing the parallel computation of tutorial 2.3 dam break. my computer is xeon 3.0 with 2 cores. so i decompose the domain with 2 parts.
but i found that the parallel computation is slow that the single core computation.
so would you please tell me what is the reason.
regards

[ngj]

Hi

The dambreak case is approx 2000 cells, thus 1000 cells per processor. A thread at the forum states that below 100000 cells per processor does not decrease the computational time, as the exchange of information increase faster than the decreased computational time. (rule of thumb).
You are well below the 'limit' of 100000 thus if you increase the number of cells you will reach a point where two processors are faster.

- Niels