User manual

The log reports that the calculation of the derivatives (the DynamicsSection) includes

413717 operations, while for the parallelization using 4 cores the longest path is 110030

operations, which means an estimated speed-up of 3.76.

The critical path is estimated to have 26131 operations, i.e., the Amdahl speed-up factor is

413717/26131 = 15.3, which indicates an upper limit of what could be obtained having

many cores and neglecting overhead.

The log then reports the structure of the parallelization obtained. First there is a sequential

part calculating 119 unknowns followed by 3 parallel layers, a sequential part and finally a

parallel layer. The log for the parallel layer 4 is opened up and it reports that there are 4

parallel sections.

Dymola supports profiling. It is activated by setting

Advanced.GenerateBlockTimers =

true

as usual. If parallelization has been activated at translation, the profiling result will

also include timing results for the sequences and the parallel layers as well as for the

individual sections. These are identified by Seq[i], Par[i] and Sec[i:j] where i and j are the

numbers given in the log above.