Pradeep,

If you have a Sandy-Bridge or Ivy-Bridge based system, and it has at
least eight cores, and those cores run at 3.4GHz, and you use the
'-openmp 8' flag (available in v5.3) and you arrange recon-all to run
the hemi stages in parallel, then recon-all completes in about 3.5
hours.  In the upcoming v6 release (expected sometime around March),
runtime in this config is just under 3 hours.  Not sure yet how much
Haswell will reduce this.

We don't actively support the GPU code as its too difficult to maintain,
but it will trim maybe 10 minutes from the total runtime over using
openmp w/8 threads.

N.

On Wed, 2014-12-17 at 15:04 +0530, pradeep mahato wrote:
> Hi Nick,
> 
> 
> Can you tell me what would be best processing time to do recon-all
> process on one brain mri image using a high end machine ( Servers ) or
> using GPU based machine.
> 
> We have  more than 1000 of brain mri images, processing all these
> images is the bottleneck now.
> 
> 
> Pradeep
> 
> 
> On Tue, Dec 16, 2014 at 6:31 AM, Nick Schmansky, MGH
> <ni...@nmr.mgh.harvard.edu> wrote:
>         Pradeep,
>         
>         Our lab will be getting a Haswell-based system soon and will
>         evaluate
>         any speed gains.  The system is a Dell PowerEdge R730 with an
>         Intel®
>         Xeon® E5‐2643 v3 3.4GHz processor.
>         
>         N.
>         
>         
>         On Mon, 2014-12-15 at 13:43 +0530, pradeep mahato wrote:
>         > Hello everyone,
>         >
>         >
>         > Have anyone used Broadwell or Haswell process for Brain MRI
>         image
>         > processing.
>         >
>         > If used can you please mention the system configuration you
>         are using.
>         > How long does it take to run recon-all process for one brain
>         MRI
>         > image.
>         >
>         >
>         >
>         > Thanking you
>         >
>         >
>         > Pradeep Kumar Mahato
>         
>         > _______________________________________________
>         > Freesurfer mailing list
>         > Freesurfer@nmr.mgh.harvard.edu
>         > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
>         
>         
>         _______________________________________________
>         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.
> 
> 
> -- 
> Pradeep Kumar Mahato


_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

Reply via email to