The Working Group on the long term preservation of research data is currently working on a complex document. The latest version of this document will be ready in July 2017. The document not only implies the assessment of what is available but also what the costs are. These are different depending on the Member States, the pan-European and the international institutions that you are considering. The cost section is the most complicated one in the document. The Working Group expects to release a new version before the e-IRG Workshop in Tallinn in October 2017.
Primeur Magazineasked Dr. González-Aranda to explain more about the objectives of the new EuroHPC Initiative.
Dr. González-Aranda started by telling that the EuroHPC Initiative is open to all the EU Member States and even the Associated Countries. The initial partners are Spain, France, Germany, Luxembourg, Italy, The Netherlands, and Portugal. (Meanwhile, Belgium has signed the EuroHPC declaration as the 8th country.) The aim is to create an ecosystem for the establishment of the exascale network, scale supercomputing with the network in Europe, and also to share our commons. EC President Juncker has declared that the European Union has to be competitive in the international arena with regard to the USA, China, Japan and other stakeholders, in order to enhance and promote the European industry in the public as well as the private sector related to HPC.
The creation of this ecosystem implies a strong commitment. Initially, it is based on a Memorandum of Understanding that was signed on March 23, 2017 in Rome, in attendance of the high-level authorities from the seven countries. The big challenge however is how to engage the community in this competitive but exciting supercomputing area, how to pave the way to exascale in two steps. First, there is the creation of two pre-exascale machines, followed by the delivery of two machines that are actually exascale. There are a lot of things to consider such as the creation of the micro-processor with European technology and the integration of the micro-processor in the European exascale machines. A challenge is also how to establish the proper procurement mechanisms to release both machines. The aim is to push the European industry, not in a closed-off fashion from the rest of the world but in a way as to integrate all the industrial components.
There is a parallel initiative, related to EuroHPC, which is IPCEI, an Important Project of Common European Interest. The IPCEI for HPC at the moment involves France, Italy, Spain, and Luxembourg but it is also open to other countries in the European Union. The connection between EuroHPC and the IPCEI is to relate this to the establishment of a commons of industry. There are a lot of standards, there are a lot of industrial initiatives in Europe and we are in a globalized free market but there has to be the guarantee of some commons from the industrial point of vue of standards. There has to be established a middleware between the high-level EuroHPC policy addressed initiative and the reality of industry. This is the main goal of IPCEI: how to engage the industry through this commons, based on the development of some case studies or pilots, including personalized medicine led by Spain; innovation 4.0 led by Italy; smart manufacturing led by France; and remote sensing led by Luxembourg.
What is positive in this initiative is that interdependently of the coordinators of each thematic area, it is open to all countries in the European Union. The important challenge in this is how to push the commons, to find the common thread throughout Europe, the common ground to bring it all under one umbrella without ignoring the interests of the individual Member States. There has to be found a balance between the high-level EuroHPC and IPCEI trying to find the interoperability mechanism for the industry, not only by engaging the big companies but also the small and medium companies.
If Europe wants to be competitive in the 21st century economy, in the knowledge economy, there is a need to integrate the big as well as the small and medium companies in a European industrial development paradigm, according to Dr. González-Aranda.
As for the users, there are not only the researchers but also the citizens. The citizens should not be in the last wagon of the train of innovations related to supercomputing and exascale. We should engage them in preparing the way to the next quantum revolution. Therefore, the public and private ecosystem has to be fostered by supporting the industry. The USA and China are moving very fast, maybe because they have more consolidated administrative and federal infrastructures. In Europe, there are 28 Member States, as well as a number of associated countries. This creates an enormous complexity. We cannot avoid that some of the technology is going to be implemented outside of Europe but at the moment, we are trying to do our best to put into value all our know how in an operational way, for instance in every e-IRG Meeting, to define the requirements and to facilitate their entrance to the ministry level and find ways to address the private companies in how to engage in these projects.
Dr. González-Aranda is one of the official EuroHPC Sherpas, together with his colleague Serge Girona from the Barcelona Supercomputing Center, appointed by the Spanish Ministry of Economy, Industry and Competitiveness to provide support for the establishment of proper governance, together with the other countries, of the global mechanisms. e-IRG is considered as one of the main stakeholders in providing advise and requirements by narrowing the gap between the e-Infrastructure networking group from the perspective of the public and private target groups and the reality of the market, and to translate and make an advisory document to these different committees. The EuroHPC has to have a structure.
If all goes according to plan, the first pre-exascale machine will be released by 2022-2023. By 2024-2025, the exascale machines will be delivered. One can make a comparison with the race to the moon. In the meantime you can imagine all the technology transfer mechanisms, the development of the micro-processor, and the involvement of the small and medium companies, which is pretty exciting. It is a going together of research, industry and the private interest to properly serve in the best possible way the public administration and the citizens in general. There are two milestones: the years 2022-2023 for the pre-exascale machines and 2024-2025 for the exascale machines.
There will also be a roadmap. To prepare, a present, there is one meeting per month. In the most recent meeting the timeline has been defined. Now, the proper call for applications to be issued by the European Commission in Horizon2020 has to be prepared, as well as the funding because the funding is essential. The EC is seriously considering to have a synergetic combination with structural funds from some of the EuroHPC countries. This is working quite well, according to the EC, in order to avoid overlappings. This includes PPI initiatives and CF funds. We are doing our best to check what is going on related to this technology and federate all this, independently from the people who are going to apply or not. The funding mechanisms have to be established as soon as possible.
The plan is to just start working and to be operational at the end of 2017 or the beginning of 2018. One hopes to have the first results in 2,5 years' time.
Dr. González-Aranda ended by saying that the importance of this is Europe, Europe, Europe. The idea is the construction of a real European Union, depending on EuroHPC, a supercomputing infrastructure in a complementary way with the European Open Science Cloud, as well as the essential role of the European e-Infrastructure Reflection Group (e-IRG) to support this process.