The ABINIT Users Mailing List ( CLOSED )

["Nuno A. G. Bandeira" <nuno.bandeira@ist.utl.pt>]

sachinjoshi9@gmail.com wrote:

> Dear ABINIT users,
> I am a new user of ABINIT. I have installed it on Windows. I tried running 
> the test file C:\ABINIT\Test_v1\t96 for alumina. But I get an output file 
> saying
> "chkinp: Checking input parameters for consistency, jdtset= 1." also no other output files are created. Can anyone please let me know what the problem is? is there any memory requirement which is not being met? Also, is it better to install ABINIT on Unix or will Windows work equally fine? 

Hi Joshi,

To answer your question: in principle there shouldn't be a noticeable 
difference between running abinit in Linux or Windows.
In practice it's a different matter however. Here's why:

1) PGI executables will afford you reasonable speed but you are limited 
to 2Gb scratch disk space for k points etc. regardless of the size on 
your disk. Some programs like Gaussian or NWChem have the ability to 
split the scratch files either automatically or by hand into 2Gb chunks 
to circumvent this problem but this, to my knowledge, is not implemented 
in Abinit. Linux PGI has not had this problem for a while.

2) g95 executables that I can provide you with are somewhat slower 
because the compiler is not yet optimised to take full advantage of your 
processor. But they run like a charm with no heap or stack (disk 
storage) limitations.

3) the development team (I presume) has little interest, or resources, 
or both in creating a fast Windows binary using, say, intel compilers 
within the .NET  Visual Studio framework.
I once tried to trick a compiler script into getting ifort to work 
within a cygwin shell but the main problems are the different file 
extension ifort uses for Windows objects (.obj instead of .o) and 
getting the right libraries to stay in the path (.lib instead of .a). 
All in all it will give you a downright headache if you attempt to do this.

Personally I use Windows binaries to run simple things (single point 
runs, band structure, reading/plotting orbitals) for full scale geometry 
optimisations I normally use a super-duper 64-bit workstation...
I think PAW might be quite appealing to use on a conventional Windows 
desktop.

Ultimately the choice is yours of course.

regards,

--
Nuno A. G. Bandeira, AMRSC
Graduate researcher and molecular sculptor
Inorganic and Theoretical Chemistry Group,
Faculty of Science
University of Lisbon - C8 building, Campo Grande,
1749-016 Lisbon,Portugal
http://cqb.fc.ul.pt/intheochem/nuno.html
Doctoral student @ IST,Lisbon
--


--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.344 / Virus Database: 267.11.9/117 - Release Date: 03-10-2005