Indeed, doing it without the motion artefact simulation seems to only use 10 go of ram. Although there was a spike in usage, making it swap, then it went down, so there seems to be some copying at some point, but it works now, so everything is fine in the end.
Thanks for the help, Samuel Le 2015-09-29 11:49, Goch, Caspar Jonas a écrit : > Hi Samuel, > > I doubt CMake is at fault... > > Generally the simulation takes a lot of RAM (I think the original simulation > was done with 32 GB). However, there is a lot of copying going on in the > background and I cannot guarantee that there is no leakage. You might want to > try doing the simulation without the head motion. That should reduce the RAM > requirements. > > Best, > > Caspar > > VON: Samuel St-Jean [mailto:samuel.st-j...@usherbrooke.ca] > GESENDET: Montag, 28. September 2015 17:45 > AN: mitk-users@lists.sourceforge.net > BETREFF: [mitk-users] Regenerating ISMRM 2015 challenge data with fiberfox > > Hello everyone, > > I downloaded the archive for re-generating the data from the challenge > website and am using the latest mitk master branch. Everything built fine on > my ubuntu 15.04 machine in release mode and I can run MITK itself without any > problem nor errors. > > Although, I tried both the command line version of Fiberfox and the graphical > view for generating the phantom, but I run out of ram at around 20% of the > ¨Generating 4-compartment diffusion-weighted signal¨ process. My computer > starts swapping out like crazy with htop showing a usage of 18.5 go of ram as > I only have 16go of ram. > > Is the generation process this ram intensive of might there be something > wrongly configured by cmake during the build process? > > Samuel > > ------------------------------------------------------------------------------ > > _______________________________________________ > mitk-users mailing list > mitk-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/mitk-users [1] Links: ------ [1] https://lists.sourceforge.net/lists/listinfo/mitk-users
------------------------------------------------------------------------------
_______________________________________________ mitk-users mailing list mitk-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/mitk-users
