[kepler-dev] [Ptolemy] Real Life expensive SDF model.

[Christopher Brooks]

Hi Daniel,

I've cc'd kepler-dev to see if they have any long running sdf models
worthy of comparison experiments with Distributed SDF.

The Rome AFRL people (two of whom are on this list) have some long running models that 
I think are SDF.  However, we can't redistribute these models.

I took a quick look at the nightly build output and nothing took
that long.
You can get the nightly build output from
http://chess.eecs.berkeley.edu/ptolemy/src/ptII/adm/logs/ptIInightly.txt

To search for time usage, I ran:
grep "ms. Memory" /tmp/ptIInightly.txt | sort -nr | head

I did not find a pure SDF model with significant time usage.

There is a parameterized SDF model at
domains/psdf/demo/BlindCommunication/BlindCommunication.xml
that takes a long time to run.

Other than that, you could look over the SDF demos and see what 
takes a long time.  Perhaps one of the graphical demos using JAI
or JMF?

_Christopher

--------

    Hi Ptolemy people!
    
    I had to miss the conference this year but i already went through some 
    of the very interested material presented this year. Hope to be able to 
    make it next time!
    
    I am looking for a SDF model that is very computationally expensive (at 
    least one of the actors consuming over few minutes). It should be some 
    real life application and something that i can use to make some 
    comparison experiments with the Distributed SDF domain. If any of you 
    can provide me or refer me to such a model please contact me.
    
    Best Regards
    Daniel
    _______________________________________________
    Ptolemy maillist  -  Ptolemy at chess.eecs.berkeley.edu
    http://chess.eecs.berkeley.edu/ptolemy/listinfo/ptolemy
--------