Hi Gregorio - I don't remember whether I actually replied to your post - apologies if I am making a mess...
Can you try and check with a command like "top" how much memory siesta consumes before it stops? Even 20 GB of RAM with a system of 2500 atoms and 400 Ry mesh cutoff doesn't seem much. And are you actually prepared for that long calculations - I have simulated a system of ~1200 atoms but I used 128 cores and 512 GB of RAM... Cheers, Bartek On Mon, 2011-11-21 at 17:03 +0100, Gregorio García Moreno wrote: > Hi, the maxima resources that I can uses are 4 proc with 20Gb of Mem. > To obtain more information y would need to check it. > This jobs are with 4proc and 8GB de memory as maxima. With 20Gb the > problems is the same. > Thanks > > > Hi Gregorio, > > > > What is your hardware? > > > > Cheers, > > Bartek > > > > > > On Mon, 2011-11-21 at 16:47 +0100, Gregorio García Moreno wrote: > >> Hi > >> I have decreased teh mesh cut off up to 100 Ry, and I obtain the same > >> problem > >> Initializing Density Matrix... > >> > >> InitMesh: MESH = 144 x 320 x 320 = 14745600 > >> InitMesh: Mesh cutoff (required, used) = 150.000 162.625 Ry > >> If I decrese the mesh cut off to match, I think that the optimization > >> couldn't be correct.... > >> > >> Thanks in advance > >> > >> > Hi.. > >> > try to run with a reduced mesh cut off and see if it works or not.. > >> > > >> > 2011/11/21 Gregorio García Moreno <[email protected]> > >> > > >> >> Hi all > >> >> I'm trying to optimize a 3D systems based on discotic molecules. The > >> >> unit > >> >> cells of around 2000-2500 atoms. > >> >> I use PBE/DZP level. > >> >> The problems is, in almost situations the work always stops here, > >> before > >> >> to start scf and geometry optimization: > >> >> > >> >> Initializing Density Matrix... > >> >> > >> >> InitMesh: MESH = 216 x 512 x 500 = 55296000 > >> >> InitMesh: Mesh cutoff (required, used) = 400.000 403.541 Ry > >> >> > >> >> I don't know which is the problem. > >> >> Maybe the computational resources such as memory? > >> >> Thanks in advance > >> >> Gregorio > >> >> > >> >> > >> >> > >> > > >> > > >> > -- > >> > Assa Aravindh.S > >> > Research Scholar > >> > IGCAR,India > >> > > >> > >> > > > > > > > > -- Groeten - Regards - Pozdrowienia, Bartek
