Editorial
After some ten years of service as director of PDC, Erwin Laure has decided to leave his post and continue his career as director of the Max Planck Computing and Data Facility in Germany. PDC and the Swedish high-performance computing (HPC) community gratefully acknowledge his service and wish to express our very best wishes in regard to his future career. It is diffcult and perhaps even unfair of me as a newcomer to summarize Erwin’s achievements in a few sentences, and there are certainly people from within the PDC organization that would be better suited to do so. But, as a long-term user of the Swedish HPC infrastructure and a previous director of the National Supercomputer Centre (NSC), I feel able to at least provide a biased view of Erwin's work at PDC. As a user and program developer, I have been challenged by his profiling PDC as the leading Swedish HPC centre for large-scale parallel computing which has served as an impetus for us to rid ourselves of code legacy issues and start over with a new software project under the name of VeloxChem as presented in the PDC Newsletter No. 2, 2019 ( www.pdc.kth.se/about/publications/pdc-newsletter-no-2-2019/veloxchem-1.945304 ). As a peer director of NSC, I felt that Erwin embodied Swedish HPC on the European arena and his engagement in European projects made PDC stand out as the most international among the Swedish supercomputer centres. With the Swedish participation in the EuroHPC project and the LUMI system, I believe that Swedish users will increasingly benefit from European HPC services in the coming future and that Erwin has made an important contribution to prepare us (HPC users and centres alike) for this change. His article titled “EuroHPC – Europe’s Path to Exascale” in the PDC Newsletter No. 1, 2019 ( www.pdc.kth.se/about/publications/pdc-newsletter-no-1-2019/eurohpc-europe-s-path-to-exascale-1.911735 ), is a worthwhile read in this regard.
While the process of finding a long-term replacement for Erwin goes on, I have been asked to function as interim director of PDC. My stake in HPC stems from research in the field of theoretical chemistry where we conduct method and program development for the purpose of simulating light–matter interactions in complex molecular systems. I came to the KTH Royal Institute of Technology in 2016 and became the new head of the Department of Theoretical Chemistry and Biology (TCB), which is now part of the School of Engineering Sciences in Chemistry, Biotechnology and Health (CBH). My HPC background includes serving as the director of NSC for four years and being a member of the PDC board for seven years, which are experiences that have prepared me with a fair insight into the operations of PDC and the ability to bridge the gap in between directorships. Equally importantly, this background has also put me in contact with several individuals of the highly skilled staff at PDC and, upon arriving, I at once felt warmly welcomed and everyone has given me fantastic support so as to make the best of the situation. Apart from issues regarding the standard operations of a supercomputer centre, the single most important task for me to attend to is the replacement of the Beskow cluster. This is an important milestone not only for PDC but for Swedish HPC as a whole and it is fortunate that the procurement project is led by a person with vast experience, namely Gert Svensson. So, in regard to points that PDC can largely control, I am comfortably guided by experienced co-workers, but there are also points that are beyond our control – such as the effects of the global pandemic on hardware production and delivery, exchange rates, and so forth – that are worrisome. The only guarantee I can provide is that all the staff members at PDC will do their utmost to bring a science-enabling system into operation (in two stages) to serve you. Stage 1 (a CPU- based cluster) is scheduled to accept SNAC allocations by July 2021 and Stage 2 (with a higher capacity) should be installed within nine months thereafter.
As impressive as we hope the capacity of PDC’s new main cluster resource will turn out, it will not reach the extreme performance levels of the resources planned by the EuroHPC Joint Undertaking. In the cover article , Anni Jakobsson and Pekka Manninen describe the LUMI system which is the Finnish-based pre-exascale EuroHPC system scheduled for operation in 2021. Sweden has a share percentage of the LUMI system and the process for Swedish researchers to get access to it is described.
In light of the global pandemic, many scientists are devoting their efforts and expertise to improve on the diffcult situation and researchers in e-sciences are no exception in this sense. In this newsletter you can read about related activities of researchers from the BioExcel and SeRC centres.
Despite the upheavals locally and globally, PDC has continued with its education and outreach programme by taking our courses, seminars and other events, like the PDC cafes , online where possible, although we were able to hold a PDC-SeRC seminar on climate modelling and a system introduction course before restrictions came into force. Regrettably, this year's PDC Summer School has needed to be postponed until 2021 and we are keeping our fingers crossed that it will be possible to have face-to-face meetings and discussions with a new batch of bright students at that time.
Patrick Norman, Interim Director PDC