ATKError: bad allocation during spin-polarized transport calculation

[xuqq]

when I do spin-polarized transport calculation，my calculation stops abnormally as follows:

# Equivalent Bulk Calculation (Initial Density for TwoProbe)
# ----------------------------------------------------------------
Traceback (most recent call last):
 

    runtime_parameters = runtime_parameters
ATKError: bad allocation
Terminated Abnormally




I  lower the number of kpoints after checking my input file several times. the results are the same.
The original xu.py is enclosed.

Thanks for all

[Anders Blom]

The error message means you are out of memory. Try running on a Linux 64-bit machine instead, since Windows 32-bit apps (like ATK 2008.10 for Windows) can only utilize the memory space below about 2 Gb, and probably the other stuff (Windows itself, etc) running on your computer is taking up a large part of that already.

Changing the k-points will not have any influence in this system, since the equivalent bulk run (which is q where you crash, and where the most memory is used) uses NX x NY x 1 k-points, and for your system NX=NY=1 already. So, the only things that could really influence the memory requirement is the mesh cut-off, since you already are running at SingleZeta basis set which, by the way, is not really a good idea. Thus, your results will probably not be very good anyway.

ATK 10.8 will be released in a 64-bit version later this month, so very nice option would be to upgrade to that (provided you have a 64-bit machine, in fact).