[Camino-users] Parallelize single voxel datasynth simulation?

[Bauman, Andrew Thomas]

Hi Matt,

Thanks for your reply, separating the intra/extracellular spins shouldn’t be required for me so I will give that a try.
Do you know if the results would be the same if I initiate separate runs of datasynth with a fewer number of spins and then average the results of all of them? For example instead of running datasynth with 100000 spins, I will run datasynth 10 times with 10000 spins each.

Thanks,
Andrew

From: Matt Budde [mailto:matt.d.budde at gmail.com]
Sent: April 1, 2019 12:21 PM
To: camino-users at www.nitrc.org
Cc: Bauman, Andrew Thomas <andrew.bauman at ubc.ca>
Subject: Re: [Camino-users] Parallelize single voxel datasynth simulation?

Hi Andrew,

If you don't require intra/extracellular spins or signals separately, you can omit this line from the plyfile:
"comment closed surface"
and the simulation should speed up considerably.

-matt


On Mon, Apr 1, 2019 at 12:50 PM Bauman, Andrew Thomas <andrew.bauman at ubc.ca<mailto:andrew.bauman at ubc.ca>> wrote:
Hi,

I am trying to run a single voxel diffusion simulation on a fairly complex mesh (500000 faces). I have noticed it seems to only run on one core and takes an extremely long time to complete (more than a week). I was wondering if there is a possibility to parallelize this command?  Can I run multiple instances of the datasynth command and combine the data afterwards?

Below is the command I am running:
datasynth –walkers 80000 –tmax 2000 –p 0.0 –voxels 1 –schemefile DWI.scheme –initial uniform –diffusivity 0.6E-9 –substrate ply –plyfile meshfile.ply > sim_output.bfloat

Thanks!
Andrew
_______________________________________________
Camino-users mailing list
Camino-users at www.nitrc.org<mailto:Camino-users at www.nitrc.org>
https://www.nitrc.org/mailman/listinfo/camino-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.nitrc.org/pipermail/camino-users/attachments/20190401/eb0bd85d/attachment.html>