For this goal our group created an additional package to the Wolfram Mathematica framework which was called “oommfUtilities” and can get micromagnetic data from each OOMMF-snapshot (.ovf or .omf-files) and it does not matter what the format of this file was: text or binary. Also our package can read the tabular text data which is hold on in .odr-files.

oommfUtilities.zip

In this archive one can find the main package, documentation, several OOMMF-snapshots and example-file for fun. Enjoy!

If you have any questions or suggestions, write to me: [encode_email email=”alexey@volkov.ca” display=”alexey@volkov.ca”]

Appl. Phys. Lett. 92, 222401 (2013), PDF.

As I know in Windows the last versions of OOMMF, like 12a4 and later (OOMMF), are basically could keep more than 1 thread and accelerate the simulations by paralleling calculations, but you may use for this multithreading Tcl/Tk.  In the same time, in Linux we have  a little more difficult task, because it’s insufficiently to download the last Tcl/Tk packages from the site — we must compile them with enable-threads function and below I will show how can we do it.

1. Download the last packages  of  Tcl and Tk from the official site Tcl/Tk and I recomend the version 8.5.13 or 8.6.0, because the version 8.4.19 sometimes gives problems.

2. When your download process was fully completed then you should unzip the packages and compile it. So go to the tcl8.X.X directory that will appear after unziping and go to the unix directory:

cd tcl8.X.X/unix

After that make the commands:

./configure --enable-threads
make
make test
make install

After your compiling will complete go to the directory tk8.X.X/unix and make same commands:

 ./configure
 make
 make install

After this you will get totaly installed tclsh on our system, in my case I find it in the /usr/local/bin-directory, so that’s why I will use below tclsh in this way, for your system you can find it by using the command:

whereis tclsh

3. Now you should download the OOMMF package from the official site and I chose for myself the new version of it (12a5), this version is basically was created for parallel-computing as one can see from the comand output:

/usr/local/bin/tclsh8.6 oommf.tcl +platform

I think it’s more user-friendly but first things first. When you download the OOMMF package unzip it, go to the OOMMF directory and make the last command that I wrote. You must get something like this:

$ /usr/local/bin/tclsh8.6 oommf.tcl +platform
oommf.tcl 1.2.0.5  info:
OOMMF release 1.2.0.5, snapshot 2012.09.28
Platform Name:		linux-x86_64
Tcl name for OS:	Linux 3.2.0-4-amd64
C++ compiler:   	/usr/bin/g++ 
 Version string:	 g++ (Debian 4.7.2-4) 4.7.2
Shell details ---
 tclsh (running): 	/usr/local/bin/tclsh8.6
                  	 --> Version 8.6.0, 64 bit, threaded
 tclsh (OOMMF): 	/usr/local/bin/tclsh8.6
                  	 --> Version 8.6.0, 64 bit, threaded
 filtersh:           	/home/alex/oommf/app/omfsh/linux-x86_64/filtersh
                  	 --> Version 8.6.0, 64 bit, threaded
 tclConfig.sh:        	/usr/local/lib/tclConfig.sh
                      	 --> Version 8.6.0
 wish (OOMMF):        	/usr/local/bin/wish8.6
                  	 --> Version 8.6.0, Tk 8.6.0, 64 bit, threaded
 tkConfig.sh:         	/usr/local/lib/tkConfig.sh
                      	 --> Tk Version 8.6.0
OOMMF threads:         	Yes
  Default thread count:	  8
  NUMA support:        	  No
OOMMF API index:       	20120928
Temp file directory: 	/tmp

As you can see from this moment we can start compiling OOMMF.

4. For OOMMF compiling make next commands:

/usr/local/bin/tclsh8.6 oommf.tcl pimake upgrade
/usr/local/bin/tclsh8.6 oommf.tcl pimake distclean
/usr/local/bin/tclsh8.6 oommf.tcl pimake

From the moment of the end of compilation you can run your parallel simulations with the command:

/usr/local/bin/tclsh8.6 oommf.tcl boxsi -kill all sample.mif  -threads N

where N – is the number of threads, by the default it set to the 4.

Enjoy the result!

For more information you could use OOMMF userguide and links: http://www.tcl.tk/doc/howto/compile.html  and http://www.ehu.es/sgi/software-de-calculo/oommf

For any questions: [encode_email email=”alexey@volkov.ca” display=”alexey@volkov.ca”]

Pillar structure for spin-polarized current

I am pleased to invite everyone to the scientific workshop, which will be held on 14th January at the Bogolyubov Institute for Theoretical Physics (14-b, Metrolohichna str.Kiev, Ukraine) and will start at 3 PM.

On this workshop I want to tell about the nature of spin-polarized current in magnetic multilayers systems, I will discuss some points of theoretical descriptions of this process, appearance of the spin-transfer torque in such systems and also I show main experimental results in this field.

For any questions: [encode_email email=”alexey@volkov.ca” display=”alexey@volkov.ca”]

Oleksii M. Volkov

Phys. Rev. B. 92, 144401 (2012), PDF.