Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Topics - esp

Pages: 1 [2] 3 4 5
16
i wanted to save the device cfg after the initial density calculation, but every time i load it it tells me there is no initial information there, and it has to start over ..

here is where i am saving after this part:

   calculator = my_DeviceHuckelCalculator(oList)
   device_configuration.setCalculator(calculator, initial_state=device_configuration)
   device_configuration.update()
   nlsave(<filename>, device_configuration, object_id="cfg")

then next time when i run i use:

device_configuration = nlread(<filename>, object_id="cfg")[0]

but it always then tells me there is nothing there ... the file is 300MB after i save the cfg above .. how do i accomplish what i am trying to do?

17
General Questions and Answers / FFT
« on: October 16, 2012, 05:34 »
I wanted to speed up calculations, and I thought i saw a post somewhere saying i could use the FFT solvers to do this ... i tried it with GNRs but it said that FFT cannot be used when you have gates and dielectrics, which i do .. is this correct?  if i can use fft solvers to speed up calcs for graphene ... can i see an example script?

18
General Questions and Answers / memory errors
« on: October 14, 2012, 20:06 »
i ran some 2 probe GNR devices just fine when they were only 10nm long, but I tried to do 20nm long and starting getting the pseudopotentials error .. I believe it has been translated to "out of memory" ... i reduced kpoints to 51 in the z direction, and i reduced number of history steps to 12, still cannot get it run with 4 nodes, 2.5GB memory per core ... this was running in a cluster with intel MPI .. i tried also instead of specifying pmem=2500mb, i tried a total memory allocation of mem=80gb (20GB per node) but then i got some abort error after some time

the funny thing is, i have a lab computer (not MPI) that i can run using mpich2, it has 24GB memory, .. it has no issues except it ran all night and got nowhere ...

it seems about 10nm long is the limit of what i can do  ... again, 10nm long device runs ok, about 2-3 hours for the first calc, then 1.5hours per IV point on a transmission calc ... anyone have suggestions given this information?

19
how do I check the status of a calculation?  I mean, if i am doing a transmission calculation, for example, what can I read from the output log to tell me how far along it is?  i dont like waiting 12 hours and not knowing if it is 10% done, 50% done ... i just have no idea ... are there any markers in the output I can search for?



20
General Questions and Answers / plotting geometry?
« on: September 24, 2012, 05:59 »
I saw in one of the tutorials, some code to export the geometry to a file, and i believe the comment said "for plotting" ... is there a way to plot or make an image export of the geometry? that would be great

21
General Questions and Answers / segmentation faults
« on: September 16, 2012, 19:02 »
I keep getting segmentation faults ... i thought it was because of an ldos calc so i tried a transmission calc .. same thing but after many hours .... do you think i should have someone check out the machine?  if it is simply that atk is using too much memory, why doesn't atk use only as much as needed, or as much is available?  why allow it to create this segmentation fault?

22
we have a license which allows N masters and M slaves ... how can I check at any given time how many are being used, and how can I specify how many of each to use on a particular job?

23
General Questions and Answers / ldos best setup
« on: September 14, 2012, 21:39 »
I need to know what is the best setup for doing lddos calculations on graphene devices ... more nodes? serial? more memory? 51 k points instead of 101?  it takes forever to run this calc and i need to speed it up ... hard to be productive ... i have a choice between different machines with different specs ..

it takes something like 12 hours for -3 to 3eV in increments of 0.2, at only one bias cfg

24
Scripts, Tutorials and Applications / transmission plot
« on: September 14, 2012, 21:24 »
is there a script for plotting the transmission spectrum as done in vnl?

25
Future Releases / cannot read file names
« on: September 13, 2012, 17:19 »
in builder after importing a cfg, there are items at the bottom, since my naming of the object is long, i cannot read them ... should have mouse over to read full object id

26
General Questions and Answers / plot colors
« on: September 13, 2012, 17:08 »
i did an LDDOS calculation, and plotted the results using scripts from quantumwise.com, and i get as attached. 

I am wondering why the colors are repeating and why there are two bars on the right.  The part of the code where the colors are defined is:
Code
import pylab
#plot the LDDOS(E, z)
pylab.xlabel('z (Angstrom)',fontsize=12,family='sans-serif')
pylab.ylabel('Energy (eV)',fontsize=12,family='sans-serif')
contour_values = numpy.linspace(minLDOS,maxLDOS,51)
pylab.contourf(X, Y, Z, contour_values)
pylab.colorbar()

attached are plots produced from the same code, but with different Vgs

27
Future Releases / programmatically create perspective views
« on: September 13, 2012, 09:26 »
as we can see our device configurations in vnl, i would like to programmatically create perspective views at arbitrary viewpoints and angles  (top, side, bottom, and perspective) and output to png .. this would help for example if i run an LDDOS and i want to line up the cfg with the calc results, ... currently i have to open vnl, open the file, set the view, ... painstaking ... was this added in the latest version? if not, would be very helpful

28
General Questions and Answers / method to reduce memory in ldos?
« on: September 8, 2012, 10:25 »
i had an ldos calc quit on a linux machine due to "memory error" .. so I am wondering, which settings can I tweak to reduce the memory usage?  it is the number of history steps that is causing it (i think i have it at 20)?  .. what can i try please?  the configuration has at least a few hundred atoms (graphene of length 12nm), so it takes over 12 hours to run, so i want to try to reduce time with trial and error

29
General Questions and Answers / runtime with asymmetric doping
« on: September 4, 2012, 21:08 »
I am wondering, I made a graphene transistor device with symmetrically doped source and drain, runtime for a transmission spectrum calculation was something like 40 minutes per Vds, Vgs point ... then I made another of same size with different doping in source and drain, and it has run i think for like 10 hours and has not completed the first point yet ... it is still running, no errors, just taking a long time .. does this sound reasonable or does this make sense to you?  Even if the symmetric case re-used calculations, still i would not think it would take longer than 2x the time ...

30
General Questions and Answers / conductance plot
« on: September 2, 2012, 22:40 »
I dropped my nc file with transmission spectrum results into the conductance analyzer in vnl, and I got:
ValueError: could not convert string to float: [vg0.0V][vds0.1V]

i saved all the results as trans[vgXX][vdsXX] ... why would it try to parse the object name?

Pages: 1 [2] 3 4 5