Charm++ parallel programming system

[Pritish Jetley <pjetley2 AT illinois.edu>]

I presume you've downloaded and compiled the code. To generate input
data sets, do "make plummer" and "make tipsyPlummer". Then, do the
following:

jetley@resilience:~/work/benchmarks/barnes$
 ./plummer 0 1238 dataset.dat
jetley@resilience:~/work/benchmarks/barnes$
 ./tipsyPlummer dataset.dat
dataset.tipsy.tbin

The first command generates a Plummer model file with 1238 particles.
The zero you see on the command line is to tell the program to write
out a file named "dataset.dat". Of course, you can give any name to
the Plummer model file.

The second command converts this file into Tipsy format.

To run, set the following:

* -in=<input Tipsy file name>
* -killat=<iterations>
* -p=<num tree pieces>
* -ppc=<max particles per tree piece>

The product of p & ppc should be about 1.2-1.5 times the number of
particles; otherwise, you will get an error message telling you that
the simulation was started with too few tree pieces; if this happens,
set -p to the value suggested in the error message.

Let me know if you run into trouble.

Thanks,
Pritish

On Wed, Sep 26, 2012 at 4:47 AM, Junchao Zhang 
<jczhang AT illinois.edu>
 wrote:
> Hello,
>   I downloaded barnes in Charm++ from
> git://charm.cs.illinois.edu/benchmarks/barnes
>   I can build it. I can also know what arguments should be added to run it
> by looking at Main.cpp
>   But I don't know what values should be chosen to get good performance, for
> example, on the Taub cluster.
>   Is there any documentation on that?
>   Thanks!
>
> --Junchao Zhang
>
>
> _______________________________________________
> charm mailing list
> charm AT cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/charm
>



-- 
Pritish Jetley
Doctoral Candidate, Computer Science
University of Illinois at Urbana-Champaign