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.


Messages - serhan

Pages: 1 2 [3] 4 5 ... 7
31
General Questions and Answers / Re: Choice of basis
« on: April 26, 2009, 16:58 »
Thanks again.

32
General Questions and Answers / Choice of basis
« on: April 26, 2009, 07:02 »
Hi,

What is the criterion for the choice of basis set type? Element type or geometry or so? To be spesific, can we use SZ sets for graphene calculations?

Cheers,
Serhan

33
General Questions and Answers / Re: St9_badalloc error
« on: April 26, 2009, 06:01 »
I'm using Ubuntu 9.0.4. However, I could not install VNL on it, it gives some library errors, that may be corrected by installing libraries. I'm using only ATK on it.

34
General Questions and Answers / Re: St9_badalloc error
« on: April 25, 2009, 16:40 »
I'm using on 64-bit Ubuntu. Does ATK automatically tries to assign swap space, or is it dependent on th OS?

Cheers,
Serhan

35
General Questions and Answers / Re: St9_badalloc error
« on: April 25, 2009, 15:06 »
Hi,

Can I use swap space if RAM is not enough?

Cheers,
Serhan

36
General Questions and Answers / Re: "Killed signal 9" error
« on: April 23, 2009, 21:16 »
Thanks.

37
General Questions and Answers / Re: "Killed signal 9" error
« on: April 23, 2009, 20:52 »
I asked this to MPI mail list and the answer is:

Code
Try the latest version of MPICH2, 1.0.8. It is hard to say what the problem might be. Could even be a bug in the application.

This may help others that gets the error.

Cheers,
Serhan

38
General Questions and Answers / "Killed signal 9" error
« on: April 23, 2009, 04:59 »
Hi all,

I'm trying to simulate a (6,6) perfect nanotube. My py file is attached. I have 4 reps in electrodes and 10 reps in central region. But, ATK (or MPI) gives the error:

Code
# ----------------------------------------------------------------
# Electrodes Calculation
# ----------------------------------------------------------------
# sc  0 : Fermi Energy =    0.00000 Ry
rank 0 in job 2  node01_40687   caused collective abort of all ranks
  exit status of rank 0: killed by signal 9

Is there  a way to overcome this problem?

Cheers,
Serhan

39
General Questions and Answers / Re: St9_badalloc error
« on: April 21, 2009, 20:39 »
Yes, it is said to be the band structure of graphene. I've taken it after some googling from the Internet.

Cheers,
Serhan

40
General Questions and Answers / Re: St9_badalloc error
« on: April 21, 2009, 03:39 »
Thank you very much :)

Cheers,
Serhan

41
General Questions and Answers / St9_badalloc error
« on: April 20, 2009, 18:38 »
Hi,

I'm trying to calculate scf of a (6,6) carbon nanotube but ATK gives the error:

Code
master@node01:~$ mpiexec -n 12 /home/master/atk-2008.10.0/bin/atk /home/master/6/6_6_nanotube_script.py > /home/master/6.dat &
[1] 10244
master@node01:~$ Traceback (most recent call last):
  File "<string>", line 916, in ?
ATKError: St9bad_alloc

I searched google, some say that this is caused from insufficient memory. My simulation includes 480 carbon atoms. I'm using LDA with 150 Ry cut-off. Each machine uses 2GB RAM with 64-bit Ubuntu and I'm using 6 PCs in parallel.

Cheers,
Serhan

42
Hi,

May be you can use I-V characteristics of a two-probe structure for calculation of phonon frequency (for channel) since the saturation occurs due to phonon scattering. However I do not know the equations exactly, you can find exact eqs in the books.

Cheers,
Serhan

43
Hi,

I've a suggestion for the speed up of ATK for the future releases. We've used an electromagnetic software and the provider offered an acceleration option using the specially designed cards using GPU (Graphical Processing Unit) for FDTD calculations. The link is as follows for the accelerator cards: http://www.remcom.com/xf7/xstream . I do not know what happens exactly inside the ATK motor (I know very complex thing occur :) :) :) ), but I thought ATK may use an accelerator solution.

Cheers,
Serhan

44
Thanks, I've posted from my friends's computer thus the user name was different on the above post...:)

Thanks again.

Cheers,
Serhan

45
Hello,

When creating nanotube as two-probe, we do not need screening layers since electrodes are the same type with scattering region. Am I right?

Cheers,
Serhan

Pages: 1 2 [3] 4 5 ... 7