[parsec-users] TBB version Fluidanimate

Christian Bienia cbienia at CS.Princeton.EDU
Thu Jul 9 14:08:10 EDT 2009


Hi Grishma,

 

There should be no skipped particles. On my machine the output of the tbb
version is as follows:

 

PARSEC Benchmark Suite Version 2.1

Loading file "in_300K.fluid"

Number of cells: 135424

Number of particles: 305809 (0 skipped)

Saving file "out.fluid"...

 

Particles are skipped because the current implementation of fluidanimate has
a limit of 16 particles per cell. If the input file specifies more than that
then the additional particles are skipped but this is not the case for our
input files. I suspect something went wrong while reading in particles from
file.

 

I cannot reproduce this error on my machine, but you might want to take a
look at the initialization code yourself which you can find in the InitSim
function. It is fairly small and almost the same for all fluidanimate
versions. This should make it easy to find the issue.

 

Please let me know if you find out anything.

 

Best,

Chris

 

 

From: parsec-users-bounces at lists.cs.princeton.edu
[mailto:parsec-users-bounces at lists.cs.princeton.edu] On Behalf Of grishma
kotecha
Sent: Thursday, July 09, 2009 1:49 PM
To: parsec-users at lists.cs.princeton.edu
Subject: [parsec-users] TBB version Fluidanimate

 




Hi 

 

I am running the TBB version of the fluidanimate. And on command line it
gives following output with input_300K.fluid file

PARSEC Benchmark Suite
Loading file "C:\Inputs\fluidanimate\in_300K-Large.fluid"
Number of cells: 135424
Number of particles: 31 (305778 skipped)
Saving file "out.fluid"...

 

If I run the pthread version with same input parameters (same inputfile)
none of the particles are getting skipped.

(the output which i get with pthread version is: 

 

PARSEC Benchmark Suite
Loading file "C:\Inputs\fluidanimate\in_300K-Large.fluid"...
Number of cells: 135424
Number of particles: 305809 (0 skipped)
Saving file "out.fluid"... 

)

Is there any reason as in why the particles are getting skipped in
'tbb-version' and not in 'serial version' and 'pthread version'?

 

Thanks, 

Grishma

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.cs.princeton.edu/pipermail/parsec-users/attachments/20090709/f575a00a/attachment.htm>


More information about the parsec-users mailing list