Hi Simone,
I ran into the same issue in the past. You need to set environment variable CUDA_VISIBLE_DEVICES to only contain the device you want to submit the recon-all to. eg. export CUDA_VISIBLE_DEVICES=0 ;recon-all -s ... -i ...... You have to balance yourself between the GPU available.
Basile
On Wed, Dec 21, 2016 at 9:48 AM, simone coscetti <simone.coscetti@pi.infn.it
wrote:
Hello FreeSurfer Developers,
I’m attempting to make some performances evaluation on recon-all using GPUs on a node with 4 nVidia K80. Submitting different instances of recon-all on 8 different subjects, every instance run on the same device (#0), compromising the efficiency of the test. Is there a way to address different instances on different devices? Thank you in advance.
FreeSurfer version: freesurfer-Linux-centos6_x86_ 64-stable-pub-v5.3.0.tar.gz cudawn9:~ # cat /etc/SuSE-release SUSE Linux Enterprise Server 11 (x86_64) VERSION = 11 PATCHLEVEL = 4 cudawn9:~ # uname -a Linux cudawn9 3.0.101-63-default #1 SMP Tue Jun 23 16:02:31 UTC 2015 (4b89d0c) x86_64 x86_64 x86_64 GNU/Linux
Simone
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.