Geeklog Site

New release of RSPt, now with manual!

Mon, 09 Nov 2009 17:53:53 +0100

Finally, a new release with bugfixes, new output and a MANUAL!! The annoying parallel bug in rspt480 is gone together with a few other minor bugs. A few plotting routines are added for you to explore but the main addition is a manual. Yes, we have finally put our heads together and managed to write a manual that hopefully makes sense. It covers the machinery behind RSPt, how to compile and run the different parallel and serial versions, the basic input (gosh, we could fill a book with that), the most important output and much more. The manual is work in progress so please let us know what is missing. The wiki will be updated accordingly but we all have our daytime jobs to take care of. Babies are born, new positions to apply for, applications for grants to be written, broken arms and all the rest. But you don't have to worry. The RSPt team is working hard. Things to look for in the next release: charge self consistent DMFT and SIC, LDA+U, orbital polarization and much more. Stay tuned!

Bug in rspt480

Sat, 12 Sep 2009 20:19:00 +0200

There is a (known) bug in rspt480! For some materials the program will seg-fault in code2p_ad_orbmom when run i k-point parallel mode with spin orbit . In serial mode or without spin orbit the code will run without any problem (that we are aware of). The program will either die or run as expected. We are working on a fix, really hard actually. Yes, really, really hard, trust us. There are a few goodies that we will throw into the distribution to compensate for this mishap. The workshop in Sulmona was very productive so keep an eye on this space for more news.

New version of RSPt!

Wed, 02 Sep 2009 16:11:00 +0200

Finally, a new version of RSPt can now be downloaded from the download section. RSPt480 has support for k-point and band parallel calculations using MPI and FFT parallel calculations using threads (DMFT only over k-points). The parameter file para.glo is gone and there is no longer any need to recompile the code for each problem size. All compiler options are collected into one file, RSPTmake.inc. The LDA part has been improved resulting in shorter execution times and a smaller memory footprint. The DMFT part has also been improved with new solvers (e.g. HIA), more functionality and higher speed. Using the DMFT module LDA PDos can now also be ml-resolved and much, much more. In the package some more (mostly) or less relevant documentation is included but still no manual. We will try to keep the wiki updated to reflect all the changes but we can't neglect our day-time jobs too much:-)

The 4th annual workshop

Thu, 26 Mar 2009 15:53:00 +0100

The 4th annual RSPt workshop will take place in beautiful Italy 6-11 September. There will be practical hands-on sessions as well as technical discussions. See the calendar for more details.

Submit your set-up!

Thu, 08 Jan 2009 11:22:29 +0100

The user base is growing and more and more people from outside of the groups and labs where RSPt was developed are now using the code. This also means that the code is run on a more heterogeneous set of computers. Even though the code is rather straight forward to install there is no autoconf functionality in the released version. Therefor you are all encouraged to submit your setup that you are using. There is a thread in the Forum under Installation where you can post: your hardware, OS, compilers and CCPARMS and FCPARMS and a short description of any changes you had to do in the Makefiles to make RSPt run on your platform of choice. /Admin

Problem with mail again

Tue, 02 Dec 2008 07:39:18 +0100

After we moved the server where we keep www.rspt.net we have had some problems with the outgoing mail. If you have registered recently but not yet received any reply, please contact me. NOTE!! You can still not use a gmail address!! /Admin

RPSt - A (Graphical) User Interface for RSPt

Mon, 15 Sep 2008 06:31:22 +0200

I would like to announce that RPSt is ready to be Alpha tested! Order now and you will get a light weight Graphical User Interface for no extra charge! For more information see the User forum. //Patrik

Post-Doc position

Mon, 14 Jul 2008 22:23:00 +0200

Ann Mattsson at the lab in Sandia is looking for a post-doc who will work on DFT functional and code development. Click on Read More below for details. I am searching for good candidates for a post doctoral appointment in DFT functional and code development. The position will emphasize DFT exchange-correlation functional development for extending the accuracy of existing methods to new application areas. The goal is to enable predictive DFT based simulations to solve challenging scientific problems within all areas of physics. DFT code development will also be a part of the work, with a focus on the all electron, full potential, LMTO code, RSPt (see http://www.rspt.net). The job is posted at our groups web pages http://www.cs.sandia.gov/materials_methods/news1.html where you can also read more about our group and what we do. You can read more about post-doctoral positions at Sandia at http://www.sandia.gov/employment/special-prog/postdocs/ The job is also posted at the Sandia Job Openings page, where a candidate also should input his application: http://www.sandia.gov/employment/career-opp/index.html The post-doc position has reference number 60918 and the posting date is 07/08/2008. Many thanks, Ann.

Don't forget the workshop!

Tue, 01 Jul 2008 07:23:00 +0200

Don't forget to register for the RSPt workshop! The workshop is the right place to meet the people behind RSPt, learn more about the project and perhaps join the group of developers. See the story below for more details.

New version of RSPt

Tue, 27 May 2008 11:01:00 +0200

There is a new version of RSPt in the download section now, rspt144. We have removed a bug in the memory management related to fftw3 and we have also (thanks to Anna) added support for orbital moments. Orbital moments are now printed in the out file when spin-orbit and spin-polarization are used. /Admin