Organizers: J. K. Dewhurst, E. K. U. Gross, and S. Sharma
1. SUMMARY
The Psi-k funded Elk-code tutorial took place at the Max Planck Institute of Microstructure Physics (MPI-Halle) in Halle, Germany, from September 3 – 7, 2018. There were a total of 14 speakers and tutors of the code and 45 students from 16 countries. We were oversubscribed and ,unfortunately, had to turn down several applicants (we received in total 61 requests for attending the Elk-tutorials).
The Elk LAPW code (http://elk.sourceforge.net/) is an electronic structure code based on the state-of-the-art full-potential linearized augmented plane-wave (LAPW) method. It was designed from the start to be a user- and developer-friendly code, allowing PhD students and post-docs to both use the code for their research as well as implement new ideas in the field of electronic structure.
The present Elk-tutorial was fourth in the series of tutorials. The previous tutorials were held in 2011, 2013 and 2015. The aims of the tutorials have been introduction to the ELK code as well as the cutting edge science and implementations in the field of electronic structure methods.
The tutorial was divided into two parts:
1. Morning sessions which were aimed at bringing together experts in the field of electronic structure methods, density functional theory (DFT), the LAPW method, superconductivity and time-dependent DFT. The talks given by the speakers were to teach the attendees the physics behind various present and future implementations in the Elk code. This session was also aimed at introducing the cutting edge topics (like warm dense matter and materials search). All the presentations will be made available on the Elk web-page (http://elk.sourceforge.net/).
,2. Evening sessions were aimed at (a) teaching attendees the most efficient method of using the Elk code and (b) discussing implementational as well as advanced level problems with running of the code. For this, 2,000 processors on the MPI computer cluster were reserved. All the participants were given accounts on this cluster. They were allowed to run the examples on this cluster for the five days of the tutorial.
2. GOAL AND IMPORTANT OUTCOMES OF THE ELK TUTORIAL
The main objective of the tutorial was not just to teach attendees to use the code, but also to introduce the cutting edge physics issues and progress in this direction in terms of implementation. In order to ensure that these cutting edge but computationally expensive tasks could be run, we opened accounts on our computer cluster in MPI-Halle for all the attendees. They were able to run in parallel the code on our cluster, which for the period of the workshop was reserved for just this purpose. Furthermore, this also allowed the attendees to experience a realistic environment of running the code on any supercomputer in the world. Several students found it chaotic in the beginning but were able to learn this very quickly. We feel this is a very important aspect of running codes and should be a part of tutorials.
The aims of the hands on part of the tutorial was to
1. introduce the basic code
2. allow trouble-shooting for advanced users and
3. introduction of the cutting edge features.
,All the objectives were met. In fact the trouble shooting part of the tutorial was especially appreciated by many students. Several of these students have explicitly requested PDFs of the lectures. These will be made available on the Elk code web-page.
On the second day we had a special session asking people for a ”wish-list”. We were given a wish-list by the students and we believe this will help us in taking the code to the next level of sophistication.
The final outcome of the Elk tutorial was that we have trained many early PhD students in using the code. They were able to ask questions about the physics behind DFT LAPW codes to the experts in the field. The more experienced participants were able to trouble shoot by asking very specific questions to the developers of the Elk code.
In a special session we discussed the future directions the Elk code must take and this also provided a forum for discussion amongst the developers of the Elk code.
TOPICS COVERED IN THE WORKSHOP
The following topics were extensively covered during the workshop
• Basics of density functional theory
• Linearized augmented plane wave method
• Basics of magnetism
• Functionals within density functional theory
• Time dependent density functional theory
• Linear response
• Many body perturbation theory
• Super-conductivity
• Phonons
• Warm dense matter
• Introduction to Elk
STUDENTS’ FEEDBACK
The students showed great interest in fundamental talks given by experts as well as in learning the code. Some of the students have shown interest in not just using but also implementing new features in the code.
Some of the students attending the workshop were experienced users of the Elk code and were able to discuss advanced level problems with the developers during the evening sessions. They also provided a list of changes made and features added to the code. This has further helped the developers in improving the code.
For students we will set up a web-page with all the talks as pdf files. This we believe will be an invaluable resource for researchers using electronic structure methods.
Several students wrote to us after the tutorial and we include few of these below.
• Sylwia Golab, AGH University, Poland In my opinion Elk Workshop is very well-organized. It is a really pleasure to be here. Topics of lecture are very interesting. I am very thankful for your patience when we have problem with the simplest thing with elk code. I did not know elk code and this workshop really allow me to understand how it works and how to use it.
• Tristan Matalla-Wagner, Germany Overall I really enjoyed the workshop and I will recommend the atten-dance to anyone who is interested in DFT calculations.
• Vladislav Yakovlev, MPI-Garching, Germany Thank you once again for the excellent workshop! The talks and the organization were at a very high level, I enjoyed many discussions with other participants, and am I now seriously considering using Elk for my future projects!
• Alyn James, Bristol University, UK The Elk workshop gave me the opportunity to learn and use many different types of calculations available within Elk in a comprehensible way. The workshop was clearly very well organised and any issue was resolved very quickly! Being there gave me the chance to talk to the developers about any issue I have experienced and discuss related work with other researchers. I would recommend this workshop to any Elk (or even non-Elk) user to get an appreciation of DFT and what Elk can offer!
• Michal Rysbki, AGH University, Poland First I want to thank you for giving me possibility to participate on this workshop. I think its great organized. The level of lectures and tutorials are very high and consistent with my phd. research. I have learned a lot. Also I am greatful for you to organized accommodation, lunches, coffees, snacks and other fancy stuff.
• Eddie Harris-Lee, Bristol University, UK The whole week was extremely well organised. In particular I arrived a day later than expected due to my flight being cancelled at the last minute – this could easily have been an issue but it was handled completely smoothly. The structure of the workshop (lectures followed by hands-on sessions) was ideal. The content was sufficiently varied that I was able to learn new things about the topics that I previously had no knowledge of, as well as those that I already knew something about. Overall I have really enjoyed the workshop, and I think that I have gained knowledge that I wouldn’t have obtained otherwise.
• Jerzy Goraus, University of Silesia, Poland Elk code is very useful project, and the workshop is also very nice. What I think could be improved in elk code are the mixing schemes (in order to obtain faster convergence) and some form of occupation analysis (e.q. Bader charges) also might be added. The greatest strength of elk in its current form is in my opinion the good documentation of the source code. Many commercial codes like e.q. wien code are documented rather poorly, which makes changes and modifications more difficult.
• Gediminas Simutis, Paul Scherrer Institute, Switzerland Just wanted to thank you for the well-organized workshop, it was very good. In particular, the speakers were excellent and – even more important – made themselves available during coffee breaks etc. for questions. For someone like me, who is an experimentalist and uses DFT occasionally, it was a great opportunity to deepen the theoretical understanding as well as learn some tricks in running the calculations. Also, the practical aspects were very well thought out – great accommodation, nice dinner and good overall atmosphere throughout the week.
• Vijay Maurya, ML Sukhadia University, India This is a great experience in the workshop and with best teachers. I attended many workshop but this is the best one. I’ll want to work with you. It’s my best chance to become better. I’ll really want come again and work with you. It’s really very good for my PhD. Thank you for this workshop.
PROGRAM OF THE ELK WORKSHOP
Monday 3-Sept.2018 (Density functional theory and LAPW)
09:10-09:30 Welcome by E. K. U. Gross
09:30-10:30 ABC of DFT (E. K. U. Gross)
10:30-10:50 Coffee break
10:50-11:50 The linearized augmented plane wave method (D. J. Singh)
11:50-12:50 EFG of DFT (E. K. U. Gross)
12:50-14:30 Lunch
14:30-15:30 Introduction to Elk (J. K. Dewhurst)
15:30-16:00 Coffee break
16:00-17:30 Exercises: Setting up the input and performing ground-state calculations (Dewhurst, Müller, Sharma)
Tuesday 4-Sept-2018 (Functionals and Magnetism)
09:00-10:00 Magnetism and spin: General overview (L. Nordström)
10:00-11:00 Functionals: New and old (L. Kronik)
11:00-11:30 Coffee break11:30-12:30 LDA+U method and non-collinear magnetism in LAPW (L. Nordström
12:30-14:30 Lunch
14:30-15:30 Magnons (N. Singh)
15:30-16:00 Coffee break
16:00-17:30 Exercises: Magnetism (N. Singh and J. Livas-Flores)
18:00-20:00 Trouble shooting + Dinner (J. K. Dewhurst, S. Sharma)
20:00-21:00 Round table: What are important future directions for the Elk code.
Wednesday 5-Sept-2018
09:00-10:00 TDDFT: Introduction (E. K. U. Gross)
10:00-11:00 TDDFT advanced (E.K.U. Gross)
11:00-11:30 Coffee break
11:30-12:30 Real time propagation in solids (P. Elliott)
12:30-14:30 Lunch
14:30-15:30 Exchange parameter from Elk (A. Jacobsson)
15:30-16:00 Coffee break
16:00-17:00 Linear response and beyond with ELK (S. Sharma, N. Singh, P. Elliott)
17:00-18:00 Trouble shooting+snacks and coffee (J. K. Dewhurst, M. Fechner)
Thursday 6-Sept-2018 (Many-body perturbation theory and material search)
09:00-10:00 MBPT: Introduction (F. Bruneval)
10:00-11:00 Cutting edge structure prediction (J. Livas-Flores)
11:00-11:30 Coffee break
11:30-12:30 Hedin equations and GW method (F. Bruneval)
12:30-14:30 Lunch
14:30-15:30 Wannier functions in Elk (A. Gerasimov)
15:30-16:30 Coffee break 16:30-17:30 Tutorial on wannier (M. Fechner, A. Gerasimov)
Friday 7-Sept-2018 (Phonons, superconductivity and advanced)
09:00-10:00 Linear response phonons in LAPW (J. K. Dewhurst)
10:00-10:45 Superconductivity overview: BCS, Eliashberg and SCDFT (A. Sanna)
10:45-11:30 Coffee break
11:30-12:30 Warm dense matter (A. Cangi)
12:30-14:30 Lunch
14:30-15:30 Practical aspects of phonons (M. Fechner)
15:00-16:00 Coffee break
16:00-18:00 Tutorial on phonons and Eliashberg (M. Fechner, A. Sanna)
LIST OF PARTICIPANTS
1. Adel Belyadi, USTHB University, Algeria
2. Basey Brock, Vanderbildt University, USA
3. Fabien Bruneval, CEA Saclay, France
4. Attila Cangi, Sandia National Laboratory, USA
5. Ramiro Cardona, Universidad Colombia, Colombia
6. Wenhen Chen, China
7. J. K. Dewhurst, MPI-Halle, Germany
8. Anna Dyrdal, MLU-Halle, Germany
9. Peter Elliott, MPI-Halle, Germany
10. Michael Fechner, MPI-Hamburg, Germany
11. Simutis Gediminas, Paul Scherrer Institute, Switzerland
12. Arsenii Gerasimov, Ural Federal University, Russia
13. Sylwia Golab, AGH University of Science and Technology, Poland
14. Jerzy Goraus, University of Silesia, Poland
15. E. K. U. Gross, MPI-Halle, Germany
16. Eddie Harris-Lee, University of Bristol, UK
17. Tobias Held, TU-Kaiserslauten, Germany
18. Fumiyuki Ishii, Kanazawa University, Japan
19. Ankit Izardar, ETH Zurich, Switzerland
20. Adam Jacobsson, Uppsala University, Sweden
21. Alyn James, Bristol University, UK
22. Harry Keen, University of Edinburgh, UK
23. Jyoti Krishna, IIT-Roorkee, India
24. Leeor Kronik, Weizmann Institute, Israel
25. Yaroslav Kavashnin, Uppsala University, Sweden
26. Yamina Lakred, University of Algeria, Algeria
27. Qizhi Li, MPI-Halle, Germany
28. Daniel Lagos, University of Bristol, UK
29. Krzysztof Mackosz, AGH University, Poland
30. Tristan Matalla-Wagner, Bielefeld University, Germany
31. Vijay Maurya, ML Sukhadia University, India
32. Tristan Müller, MPI-Halle, Germany
33. Lars Nordström, Uppsala University, Sweden
34. Camilla Pellegrini, MPI-Halle, Germany
35. Luiz Nunes De Olivera, University Sao Paulo, Brazil
36. Inigo Roberdo, Donostia International Physics Center, Spain
37. Matheus Rodriguez-Alvarez, MLU-Halle, Germany
38. Michal Rysbki, AGH University Technology, Poland
39. Malte Sachs, Philipps University Marburg, Germany
40. Antonio Sanna, MPI-Halle, Germany
41. Tom Sonderson, University of Bristol, UK
42. Philippe Scheid, LPCT, France
43. Sangeeta Sharma, MPI-Halle, Germany
44. Yasushi Shinohara, University of Tokyo, Japan
45. Vishal Shokeen, MBI-Berlin, Germany
46. David Singh, University of Missouri, USA
47. Nisha Singh, MPI-Halle, Germany
48. Yasumitsu Suzuki, Tokyo University of Science, Japan
49. Paul Tanmoy, Technion-IIT, Israel
50. V. H. Tran, Polish Academy of Science, Poland
51. Haseen Ullah Jan, University of Bohemia, Czech Republic
52. Chung-Yu Wang, MPI-Halle, Germany
53. Sebastian Weber, TU-Kaiserslauten, Germany
54. Vladislav S. Yaroslav, MPI-Garching, Germany
55. Mingqi Yan, Sourthen University of Science and Technology, China
56. Jia-Yue Yang, RWTH Aachen, Germany
57. Angelica Zacarias, MPI-Halle, Germany
58. Felix Willems, MBI-Berlin, Germany
59. Celso Rego, MPI-Halle, Germany