Editorial

Stepping-up Europe’s efforts to build the European HPC ecosystem

The public-private-partnership between the ETP4HPC Association and the European Commission exists since 2014 and its numerous productive interactions and joint activities since then are now well-known to all. In particular, in the previous period the ETP4HPC was instrumental providing updates to the Strategic Research Agenda, an essential contribution for the long term planning and implementation of the European HPC Strategy. This materialized in the publication of the different Horizon 2020 Work Programs with HPC-related Calls in the Excellent Science (Future and Emerging Technologies), Leadership for Industrial Technologies (LEIT) and e-Infrastructure pillars, with a total funding of research and innovation actions in HPC of more than 332 M€ in the period 2014-2017. The impact of the Strategic Research Agenda however spreads beyond the members of the ETP4HPC, as many other participants subscribe to its objectives and actively participate in the implementation of the European HPC policy. It goes without saying that the ETP4HPC has become a key player for achieving the objectives of the European HPC Strategy as outlined in the Communication a “European Cloud Initiative”.

Can we do more to reach the ambitious objectives of the European HPC Strategy?

High performance computing is of a strategic importance for science, industry and the public sector. From our numerous interactions with the HPC community last year, culminating in the public consultation on the implementation of EuroHPC2, I conclude that the expert reader is well informed about the objectives of the European HPC strategy. I am grateful for the engagement of the ETP4HPC and its members, contributing to the adoption by the European Commission on 11th January 2018 of its proposal for a Regulation establishing the EuroHPC Joint Undertaking. According to the Commission’s assessment, which was overwhelmingly confirmed by the HPC constituency, the steps undertaken since 2012 and the available implementation model are insufficient to achieve the ambitious objective to develop in Europe a full HPC ecosystem, at par with its global competitors and in the same timeframe. A Joint Undertaking is considered the best available instrument to acquire leadership-class supercomputers securing Europe’s own independent HPC technology and system supply and stay competitive in the development of scientific and industrial HPC applications. The Joint Undertaking is expected to overcome the fragmentation of European and national programs, funding sources and instruments to reach the objectives of the strategy. This does however no mean that we will throw overboard the things that worked well and the ETP4HPC belongs to the latter. There is a key contribution to the realisation of the Joint Undertaking to be made. There is an active participation of the ETP4HPC to be played. I am looking forward to the upcoming discussions on where and how to capitalize on the strengths, the expertise and the structuring role of the ETP4HPC to implement the activities of the Joint Undertaking. The Joint Undertaking is an instrument to better link the R&D programs and its results to the acquisition and operation of HPC machines. The expectation is that this will also foster and reinvigorate an European HPC supply industry creating lead markets and making HPC resources and applications accessible to a wider user base, notably SMEs.

It will also lead to new and more powerful applications, keeping Europe at the forefront of scientific excellence and maintaining the competitiveness of the industrial users. Therefore, in the proposed Regulation on an EuroHPC Joint Undertaking, the ETP4HPC Association is included as a member.

Editorial

Boosting the European HPC ecosystem

EuroHPC: Building a world-class supercomputers infrastructure

ETP4HPC is committed to developing a European world-class HPC technology value chain. An important vehicle for this development are the Research and Innovation projects funded by the European Commission. In 2014, the HPC contractual Public-Private Partnership (cPPP) was signed with the European Commission and ETP4HPC has been contributing to the definition of the Horizon 2020 HPC Work Programme. The new EuroHPC initiative launched in Rome in March 2017 is providing a new impetus to the European HPC ecosystem. It has now been signed by 14 Members States, and more are considering joining in. ETP4HPC fully supports the proposition to organise EuroHPC as a Joint Undertaking.

EuroHPC will be operational in 2019; it will manage the acquisition and operation of world-class supercomputing machines. This legal structure will make it possible to merge multiple source of funding in order to provide the European research and science community with the supercomputing capacity they need.

ETP4HPC will play a pivotal role in the EuroHPC Joint Undertaking which will also organise and coordinate the Research and Innovation HPC programme. This should replace the current cPPP framework. ETP4HPC will be one of the private members together with the Big Data Value Association (BDVA) and will be part of the Research and Innovation Advisory Group.

Contributing to the Horizon 2020 HPC Work Programme

The different cPPPs active under the Horizon 2020 Work Programme for the period 2014-17 were evaluated by the EC mid-2017. Regarding our HPC cPPP, it was encouraging to receive quite a positive feedback. We are now looking forward to the coming 2018-20 Work Programme (WP), and we have provided the EC with our propositions early-2017. We are pleased to see these are taken into account and will materialise for example into the Extreme Scale Demonstrators, a concept which we promoted.

The HPC Strategic Research Agenda is maintained by the ETP4HPC. The third version was compiled thanks to our 230+ working group members and was released in October 2017. The SRA-3 will serve as a reference for the different 2018-20 WP project calls.

ETP4HPC pays a lot of attention to the needs of HPC users, in particular the Big Data community. Together with BDVA we organised joint workshops to improve a common technical language and synchronise our research models. The currently opened common calls for proposals reflect this effort. This convergence is also the main topic of interest of the Big Data and Extreme-Scale Computing BDEC international initiative. Finally, we are seriously investigating how the Artificial Intelligence (AI) can improve HPC applications and how HPC technology might accelerate AI.

The ETP4HPC association

The continual growth in ETP4HPC’s membership (88 members at the end of 2017, up from 80 in 2016) is a good indication of our association’s vitality. ETP4HPC will continue to work closely with the HPC ecosystem and strengthen the links between all actors of the HPC value chain. As in the past years, we will keep putting an emphasis on SME sector and our SME Working Group will continue strengthening this vital part of our ecosystem. We are looking forward to expanding our collaboration with other European associations, in particular HiPEAC and BDVA.

We are thankful to the EC for its support to our activity through the EXDCI Coordination and Support Action which will be continued with EXDCI-2.

In the coming years, ETP4HPC plans to contribute actively to the European HPC ecosystems and we intend to play a major role in the EuroHPC Joint Undertaking.

Our association

Our organisation

The European Technology Platform in the area of High-Performance Computing, ETP4HPC, was incorporated as a Dutch association in June 2012. We are an industry-led think tank comprising of European HPC technology stakeholders: technology suppliers, research centres and end users. Our goals are:

  • To build a European world-class HPC technology value chain that will be globally competitive
  • To achieve a critical mass of convergent resources in order to increase the competitiveness of European HPC suppliers and solutions
  • To leverage the transformative power of HPC in order to boost European competitiveness in science and business
  • To expand the HPC user base, especially SMEs (through facilitating access to HPC resources and technologies) and to open the possibilities for SMEs to participate in the provision of competitive HPC technology solutions
  • To facilitate the provision of innovative solutions to tackle grand societal challenges in Europe such as climate change, better healthcare, predicting and managing large-scale catastrophes and energy efficiency
  • To foster international cooperation in research and industry

Our organisation

MANAGED BY THE STEERING BOARD

The Steering Board currently in office was appointed by the members at the Annual General Assembly on September 29th 2015, and will remain in office until approval of the financial statements for the year 2017 and the election of the new Steering Board at the 2018 General Assembly. In 2017, the 15 members of the Steering Board elected were:

  • For the research centres (5 seats): BSC, CEA , Cineca, Fraunhofer, Forschungszentrum Jülich
  • For the European SMEs (3 seats): ClusterVision, Megware, ParTec
  • For the European controlled corporations (4 seats): Atos/Bull, Eurotech, Seagate, ARM
  • Fort he International companies with R&D in Europe (2 seats): IBM, Intel
ETP4HPC Steering Board and Office (May 2017)

List of 2017 Steering Board Meetings

DATE TYPE OF MEETING MEETING PLACE HOSTING ENTITY
February 13th 2017 Steering Board Meeting Audio
March 21st 2017 Steering Board Meeting Munich, Germany Hosted by IBM
March 21st 207 7th General Assembly Munich, Germany Hosted by IBM
April 25th 2017 Steering Board Meeting Audio
May 19th 2017 Steering Board Meeting Barcelona, Spain Held during the European HPC Summit Week, UPC
June 21st 2017 Steering Board Meeting Frankfurt, Germany Held at ISC 2017
September 7th 2017 Steering Board Meeting Barcelona, Spain Co-located with ACM Europe Conference, UPC
October 25th 2017 Steering Board Meeting Brussels, Belguim Hosted at the EC’s office at Beaulieu 33
December 13th 2017 Steering Board Meeting Munich, Germany Hosted by ParTec

APPOINTED BY THE STEERING BOARD

  • Chairman: Atos/Bull - Jean-Pierre Panziera
  • Vice-chair Research: CEA - Jean Gonnord
  • Vice-chair Industry: Seagate - Sai Narasimhamurthy
  • Secretary: ParTec - Hugo Falter
  • Treasurer: ClusterVision - Frank van de Hout

ETP4HPC Office

ETP4HPC has virtual offices based in several European countries which are supported by different members. The Office is in charge of editorial work for the Strategic Research Agenda (SRA), general support for communication and administration, and more generally, all executive tasks under the Steering Board supervision. Today’s office branches are supported by:

  • France: CEA, Teratec
  • Germany: IBM, Partec
  • Spain: BSC
  • The Netherlands: SURFsara

Repartition of our members by statue and by profile

Overview by countries: 86 members in 21 countries

Our association

Our members

Our association

2017 new members

We welcomed 11 organisations as new members in 2017. Amongst the new members we have six SMEs, two large companies and three research centres, covering a wide range of different domains within the HPC value chain.

2CRSI

2CRSI designs and manufactures storage systems, servers for high-performance computing (HPC) and customized IT appliances. We are a new class of hardware vendor that does a true differentiation, providing innovation and optimization for performance, density and efficiency. We provide solutions for data centers, big data and cloud industries, independent software vendors, universities and other businesses which need high-quality, efficient and reliable storage and computing.

2CRSI DNA is innovation in computing to provide new and different solutions. With strong partnerships with Intel (as HPC data centre specialists), NVIDIA (as accelerated computing and Deep Learning solution providers), Western Digital and with a complete team of accomplished engineers and PhDs (20% of the staff) we have developed a true innovation dynamic.

In 2017, 2CRSI has launched a new IT infrastructure dedicated to virtualization and high performance computing.

This new solution, an autonomous ecosystem optimized for large-scale computing, consists of rack, servers and shared parts. It achieves the highest performance for computing-intensive applications such as simulation, Artificial Intelligence and also virtualisation for: cloud gaming, CAD and 3D rendering.

For this new solution, 2CRSI has been inspired by the Open Compute Project (OCP), a community whose members are major players in IT, which aims to make data centre designs more efficient and user friendly. Thanks to its expertise, 2CRSI has combined the advantages of OCP with all the intelligence, power and flexibility of 2CRSI solutions.

The benefits of this solution are many: it improves energy efficiency with up to 24% reduction in power consumption compared to traditional systems and reduces the total operating cost of the data centre. Shared power and intelligent cooling management combined with easy maintenance make this solution essential for data centres.

This infrastructure is already used by Blade Group for their innovating Cloud Gaming service which allows anybody to rent and use a powerful computer in the Cloud.

Asperitas

With Immersed Computing® by Asperitas HPC environments can benefit from reduced energy footprint, reduced floorspace, and reduced requirements for cooling installations and other facilities. Far less infrastructure is needed than any other liquid installation, saving energy & costs on all levels of datacentre operations. Combined with optimised IT hardware Immersed Computing would be the most sustainable and efficient solution today. Ensuring the highest possible efficiency in availability, energy reduction and reuse, while increasing capacity.

Asperitas is a Dutch cleantech company focused on enabling green datacentres for high density environments and emerging technologies with Immersed Computing.

Asperitas launched Immersed Computing in March 2017. The first solution, the AIC24, is a fully integrated and modular solution for cloud, edge and HPC datacentres, enabling to be more efficient, sustainable and flexible. The Asperitas AIC24 is a closed system and the first water-cooled oil-immersion system which relies on natural convection for circulation of a dielectric liquid. This results in a fully self-contained and Plug and Play modular system. The AIC24 needs far less infrastructure than any other liquid installation, saving energy and costs on all levels of datacentre operations. Combined with optimised IT hardware it would possibly make the AIC24 the most sustainable and efficient solution available for IT environments today. Ensuring the highest possible efficiency in availability, energy reduction and reuse, while increasing capacity.

Asperitas has been working with partners like Vienna Scientific Cluster, ClusterVision and Supermicro since the development phase of Immersed Computing.

Since the introduction of Immersed Computing Asperitas has been nominated for several awards from the data centre, start-up and cleantech industry and has received the prestigious cleantech innovation Piet de Jong Award from the Dutch TNO institute.

ESI Group

ESI Group is a leading innovator in Virtual Prototyping software and services. Specialist in material physics, ESI has developed a unique proficiency in helping industrial manufacturers replace physical prototypes by virtual prototypes, allowing them to virtually manufacture, assemble, test and pre-certify their future products. The creation of a Hybrid Twin™, leveraging simulation, physics and data analytics, enables manufacturers to deliver smarter and connected products, to predict product performance and to anticipate maintenance needs.

From inception forty years ago, Product Lifecycle Management (PLM) traditional solutions have relied on the exponential power of High Performance Computing (HPC) to develop Products right the first time, initially by Computation, then by Simulation, and finally by “as built as tested” Virtual Prototyping. This has been done for the Product in its new state, without much attention for its real life once in the user’s hands.

Thanks to the in-life connectivity of the Product with the IoT and Big Data, the Hybrid TwinTM can now offer a powerful alternative to represent the aged Product in its ‘in-Service’ conditions, i.e. as it is used and integrated in its specific operational environment. This disruptive solution enables a continuous feedback to ensure optimal usage and performance during the Product entire lifecycle up to ultimate withdrawal.

Thus, the new Hybrid TwinTM solutions enable companies to virtually provide predictive maintenance and optimize assisted operation of the Product, by connecting information of the present (IoT) and of the past (Big Data), and combining with interpretations (I.A.) of the possible future (updated and extrapolated Virtual Prototype). These disruptive solutions are tackling the key challenges of Industry 4.0, while also allowing the development of new materials and manufacturing processes which will create the innovative and competitive Products of tomorrow (ex: lightweight, electric, assisted or autonomous vehicle, smart materials, etc.).

HPCNow!

HPCNow! offers a wide range of services in all our activity areas, including consulting, system installation and administration, technical and scientific support, and training. Our service portfolio is specifically designed to help our customers meet all their needs during the life cycle of their computing and storage systems. Due to our high degree of specialization and extensive experience, we offer extremely high-level management and support services, and the support cases are directly assigned to the team with the abilities to address them.

HPCNow! provides its customers with technology and solutions to address complex issues in the field of high performance computing, data storage, and Big Data systems. HPCNow! can offer experts in services that span multiple areas, including cluster design and installation, supercomputer administration and deployment, and optimization of distributed file systems. Our company also offers training and consulting services for public and private companies in science and engineering sectors. Additionally, we contribute to the development of several open source projects related to HPC. HPCNow! provides its customers with the best solutions, getting the most out of their systems and maximizing the investment made.

We stand out from our competitors by offering our extensive knowledge and expertise in HPC. We have worked and collaborated in this area with many companies and institutions in several sectors: pharmaceutical, biotechnological, chemical, meteorological, automotive, aeronautical, banking... Our experience also allows us to extend the range of our services to the field of Big Data systems. The Big Data paradigm can deal with extremely large volumes of information quickly and efficiently, making it very attractive for HPDA applications (High Performance Data Analysis). These applications constitute the point of convergence between HPC and Big Data models, as they can run tasks with such large data volume and algorithmic complexity that they require the use of HPC resources.

Iceotope Technologies Limited

Iceotope is a UK based technology company specialising in liquid cooling technology for computing and electronics. Iceotope have been developing liquid cooling technology since 2005. Iceotope is VC backed, with strategic corporate sponsors including Schneider Electric and Solvay. Our technical experts look beyond conventional methods, the result is an innovation that delivers real efficiency, making systems smaller, reliable and energy-smart. we offer a revolution in energy-efficient technology: reducing the cost of computing in every sense.

Iceotope is the global technology leader in liquid cooling technology. Stepping out from standard thinking, we deliver innovation that provides cutting edge computing in less space, with less noise, lower energy consumption and considerably lower cost by delivering cooling technology for products that is both innovative and familiar. We cool all components with engineered coolants to eliminate airflow and maximise cooling efficiency. This way, we get all of the efficiency benefits made possible by immersion cooling but our products look and feel like traditional, air-cooled infrastructure.

In 2016, Iceotope was chosen, through a competitive process, as the cooling technology for the European Union’s Horizon 2020 EuroEXA project to deliver exascale computing. Iceotope’s cooling technology allows for the density and scale necessary for exascale computing and will soon be enabling partners to push the boundaries in cosmology, astrophysics, AI simulation, medicine and nuclear fusion.

Iceotope intend to exploit cooling technologies with key industry partners, including corporate sponsor of Iceotope, Schneider Electric, and aim to find applications for the technologies in Edge Computing, HPC and Cloud Computing.

IFP Energies nouvelles (IFPEN)

IFP Energies nouvelles is a major research and training player in the fields of energy, transport and the environment. To attain its objectives, IFPEN uses high performance computing simulators to i) propose renewable sources, alternatives to oil and gas solutions like waves and offshore wind energies ii) develop environmentally-friendly solutions in the field of transport and iii) propose technologies that meet the demand for energy by improving energy efficiency and reducing the environmental impact.

At IFPEN, from research to industry, technological innovation is central to all its activities, structured around three strategic priorities: sustainable mobility, new energies and responsible oil and gas.

IFPEN develops and co-develops several high performance computing applications, which contribute to the achievement of its strategic priorities. The objective is to be able to simulate more and more complex phenomena in a reasonable time for accelerating design, development and validation of technological innovations. The contributions of IFPEN cover from fundamental research to its validation on  real-world problems, applications and data  in partnership with academic and industrials end users.

Tackling strategic priorities mentioned above has been structured into 9 interdisciplinary scientific challenges on which IFPEN focus fundamental research efforts, among them are the performance of computing codes and big data flows. On these particular topics, researches are conducted in the fields of : i) Data compression and visualization, ii)  Programming model interoperability and composability, iii) Numerical methods and algorithms enabling extreme scalability and iv) Mathematical support for data placement and data movement minimization.

Since 2006, IFPEN have been working on the multi-core architectures accelerated or not by the GPUs. The IFPEN on-going research is focused on the many-core architectures.

NEC

NEC is a leading provider of HPC solutions, focusing on sustained performance for real-life scientific and engineering applications. NEC delivers HPC technology and professional services to industry and academia. Linux-based HPC clusters as well as our high-end vector systems meet the different needs of different customers in the most flexible way. Energy-efficiency is one of the key design objectives, addressed by advanced cooling technologies or by the high-bandwidth vector-architecture, which delivers unprecedented efficiency on real world code.

NEC SX Vector Technology for Highest Sustained Performance

NEC understands and deploys every technology that is prevalent in the HPC world today. But NEC goes beyond: NEC has been a provider of vector supercomputers since more than 30 years now. Our SX-Aurora TSUBASA vector engine provides a breakthrough in performance, combined with excellent ratio of performance per power consumption combined. These systems will be closely integrated into a standard LINUX-environment, and will flexibly allow to run applications either on the vector CPU or on standard x86-CPUs, allowing scientists and engineers to select the optimal platform for each job.

NEC HPC Storage: LXFS, LXFS-Z, GxFS

Parallel filesystems are the technology of choice for HPC storage solutions. NEC’s most recent solution, called NEC LxFS-z and NEC BxFS-z, incorporates a ZFS-based Lustre or BeeGFS appliance, thus combining highest performance and scalability and data safety and integrity. NEC’s development team has full control over this product and will continue to add functionality and enhancements. For enterprise demands, the NEC GxFS Storage Appliance is powered by IBM Spectrum Scale, an extremely reliable and mature HPC storage solution. It provides a wide range of features which allow operation in HPC environments as well as in business-critical enterprise environments. The NEC GxFS Storage Appliance includes a feature-rich toolset for installation, administration and operation while the system is highly customizable to allow integration into existing environments.

PRO DESIGN Electronic GmbH

PRO DESIGN is an expert in FPGA board design and manufacturing. Technology and market leaders worldwide are using our products. We would like to extend our product portfolio for HPC and are looking for new collaborative projects where FPGA-based platforms are used.

PRO DESIGN is a SME with about 95 employees. Our scope of business includes Electronic Engineering and Manufacturing Services (E²MS) and FPGA-based HPC and (ASIC) prototyping systems – product family proFPGA: www.profpga-hpc.com and www.profpga.com.

PRO DESIGN is a worldwide recognized expert in FPGAs. We have sold and supported more than 1.000 proFPGA systems and thousands of custom FPGA boards around the globe.

PRO DESIGN is an active member in the EU-funded MANGO project (www.mango-project.eu). Our main focus is on FPGA-based hardware / PCB development and production including bring-up and test. In addition, we can offer software and HDL development. We are looking for new projects and product developments in the field of FPGA-based HPC.

Heidelberg University Computing Centre

Heidelberg University Computing Centre is the central IT service provider of Heidelberg University, Germany. Our users are the university’s students, scientists and staff, to whom we provide the full spectrum of information and communication services and support. Through membership in ETP4HPC we aim to establish the link between our regional user communities and the European community. Our primary goal is to provide low barrier access and support for state-of-the-art HPC technology including our on-premise HPC cluster and remote resources on the regional to European level.

As a member of the HPC federation of the German federal state of Baden-Wuerttemberg, we operate the HPC cluster “bwForCluster MLS&WISO” which entered the June 2015 TOP500 list. Furthermore, we run the accompanying HPC service and support centre with special emphasis on molecular life sciences, economics and social sciences. Our mission is to engage with the scientific and industrial community to promote and foster uptake of HPC technology across all disciplines. We are committed to provide leading edge HPC technology to users from all disciplines, to familiarize them with HPC for their benefit in all kinds of research and innovation, to enable our users to tackle grand challenges in research and society by leveraging powerful HPC resources, and to cooperate with scientific and industrial communities from the regional to the European level.

Our research and innovation activities include energy-aware HPC, remote visualization, and HPC technology integration with related fields such as big data storage and analysis or cloud infrastructures.

Faculty of Electrical Engineering and Computing University of Zagreb

Faculty of Electrical Engineering and Computing University of Zagreb with it’s HPC Architecture and Application Research Center is strong competence center and R&D institution in the southeast Europe region. Active participation in some of the leading EU HPC projects and expertise in domains of processor architecture, high efficiency optimizations, co-design, implementation of compute-intensive algorithms and HPC applications has enabled creation of strong R&D links with various industry/academic partners and collaborative research activities.

The University of Zagreb, Croatia (1669) is the oldest and biggest university in South-Eastern Europe with 29 faculties and three art academies and over 50,000 full-time undergraduate and postgraduate students. Faculty of Electrical Engineering and Computing is the leading Croatian and regional academic and R&D institution in the field of electrical engineering and ICT and is constituted of 12 departments, comprises more than 140 professors and 230 teaching and research assistants and has around 4500 bacc/master students and 500 PhD students.

FER HPC Architecture and Application Research Center (http://hpc.fer.hr/en/hpc) is focused on HPC R&D of high-performance, energy-efficient and application-specific computing systems. The research targets domains of processor architecture and high efficiency optimizations, co-design and implementation of various compute-intensive algorithm domains and HPC applications. Our special interest is in imaging/video algorithms on HPC platforms with processing executed on standard CPUs, GPUs, FPGAs and special HW accelerators. The Center has experience in design of numerical and algorithmic HW/SW accelerators as well as applications in e-health/medical imaging domain.

HPC R&D projects are important activity of our group so we are partners of H2020 FET-HPC project “MANGO - exploring Manycore Architectures for Next-GeneratiOn HPC Systems” and also partner of European Processor Initiative consortia.

Verne Global

Verne Global’s new high performance computing service (hpcDIRECT) is a powerful, agile and efficient HPCaaS platform. Built to address today’s intense compute requirements, hpcDIRECT provides a fully scalable, bare metal service with the ability to rapidly and securely deliver the full performance of HPC servers.

hpcDIRECT is accessible via a range of options, from incremental additions to augment existing high performance computing, to supporting massive processing requirements with petaflops of compute. This flexibility makes it an ideal solution for applications such as computer-aided engineering, genomic sequencing, molecular modelling, grid computing, artificial intelligence and machine learning.

hpcDIRECT is available with no upfront charges and can be provisioned rapidly to the size and configuration needed. hpcDIRECT clusters are built using the latest architectures available including Intel’s Xeon (Skylake) processors, and fast intercore connectivity using Mellanox Infiniband and Ethernet networks, with storage and memory options to suit each customer’s needs.

Since Verne Global began operations in early 2012, the company has been at the forefront of data center design and technology, bringing new, innovative thinking and infrastructure to the industry. hpcDIRECT is the latest product in this cycle, and is perfectly optimised for companies operating high performance and intensive computing across the world’s most advanced industries.

Our association

Reports from the Working Groups

SME Working Group

ETP4HPC recognises the value of SME participation in the European HPC Value Chain. The development of HPC technologies in Europe should provide a number of SME with an opportunity to become globally competitive players in this market. European SMEs are capable of achieving success in this process; however, SMEs and start-up face some particular hurdles. The goal of this Working Group is to federate the SMEs in HPC and to act as intermediary towards the European Commission to overcome those obstacles. To this end, the Working Group recommended to the EC specific support actions for SMEs. Moreover, in the light of the upcoming 9th Framework Programme and the EuroHPC initiative, the SME WG is currently working with its members and with the support of the EC on a strategy for SMEs. This new dynamic in 2017 – supported also by EXDCI - gave rise to a meeting amongst the SME WG in February 2017 in Haarlem (Netherlands), as well as a meeting with Dr. Kalbe in June 2017. These efforts will be pursued in 2018.

Contacts: Hugo Falter, ParTec
Frank van der Hout, ClusterVision

Industrial Contact Group (ICG)

The dialogue between HPC technology providers and HPC users has been at the heart of ETP4HPC since its beginning, for example by involving the industrial end-users in the reflexions for the Strategic Research Agenda (SRA). The goal of the Industry Contact Group is to go a step further by establishing a permanent representation of HPC users from various domains within ETP4HPC.

We are convinced that the early and upfront involvement of HPC users is beneficial for the technology providers and system designers on one hand and HPC users on the other: It helps HPC technology providers to better understand the market and their client’s needs. And the HPC users can expect that future HPC solution portfolio will address their businesses needs. To encourage this co-design approach for the upcoming Extreme scale Demonstrators, we brought together – via a workshop at ISC17 – industrial users and technology providers.

Maximising the benefit of HPC technology for European HPC user community is a key mission of ETP4HPC, and the ICG contributes to this goal via this continuous dialogue with the HPC users.

Contact: Maike Gilliot, ETP4HPC Office

HPC Software Working Group

The key motivation for the HPC Software Working Group is to increase the long-term impact of software that is developed in the European Union, e.g. as part of FP7 and H2020 projects. In this context, software is not limited to scientific applications, but includes system level software and libraries. In order to increase its impact, HPC software should be highly scalable, well designed and engineered, and sustainable.

Key topics that are being addressed are scalability and efficiency (how well are resources being utilised), software design and engineering (what does software development for massive parallelism entail), and sustainability (how can software be made useful, reusable and discoverable). As the current tranche of FETPHC projects enter their latter stage in 2018, the Working Group will survey the European HPC Software landscape to gain an understanding of how well European software addresses each of these topics, and where the major strengths and weaknesses lie.

Contact: Michèle Weiland and Mark Parsons, EPCC

Energy Efficiency Working Group

Energy efficiency is one of the most pressing issues for HPC in general and particularly on the path to Exascale. It has to be addressed at multiple levels: applications, system software, hardware, and infrastructure. The aim of this working group is to create awareness about latest research and developments on energy efficiency in HPC, share experiences, and to promote best practices on the design and use of HPC systems and infrastructures among ETP4HPC members. The working group is led by LRZ and open to all interested members.

Contact: Michael Ott, LRZ

Our main activities

Strategic Research Agenda (SRA) – paving the way towards leadership in HPC technology

The Strategic Research Agenda (SRA) is our key deliverable, constituting a roadmap towards a European globally competitive HPC technology provision.

Working session of the SRA Working Group leaders at IBM Rüschlikon (July 2017)

This is our third Strategic Research Agenda (SRA). We maintain a valid European HPC technology roadmap in place at any time as one of the key objectives of our association. As before, the process of writing the SRA was an open one, with all the members of ETP4HPC having an opportunity to contribute their expertise. This is the result of the collective work of nearly 230 experts in eight technical working groups as well as some other non-technical task forces.

The role of the SRA

This SRA outlines the European research priorities in the area of HPC technology provisioning. The provision of HPC technology is viewed as one of the three pillars of European HPC, alongside infrastructure and application expertise. The importance of HPC is now widely recognised and European HPC has a good momentum due to the investments and programmes, such as the contractual Public Private Partnership in HPC (HPC cPPP). Europe is improving its position to compete with other regions in the area of HPC technology provision and matches its weight in the consumption of HPC systems. The challenges in the area “extreme compute and extreme data” received a special attention in this SRA.

ETP4HPC, as part of the European HPC ecosystem, issues its SRA, which is used by the EC as a recommendation in formulating its research programme. In a following step, the EC’s calls for proposals (related to the HPC cPPP) are announced: The SRA is expected to be used as guideline in the process of creating and assessing projects within the EC’s HPC technology Work Programme (WP). In particular, any project submitted within WP18-20 (the last part of the EC’s Horizon2020 Framework Programme) should cover milestones included in the third release of our SRA.

SRA’S ROLE

The SRA influences the process of defining EC calls for proposals in the area of HPC technology. The guidelines included in the SRA are expected to be used as a reference for the call objectives and assessment criteria for project proposals submitted.

The SRA Model

The “Multi-dimensional HPC Vision” defined in SRA 1 (2013) remains till now the foundation of the SRA structure as a proven approach in dealing with multiple facets of HPC technology.

The “north” dimension of this model points at the major areas of R&D of new technologies able to offer more competitive and innovative HPC systems for a broad HPC market, and the “east” side at enhancing these technologies with the right characteristics to address the extreme-scale requirements.

The third element – the “south” direction - is the current trend of developing new HPC applications. Besides traditional HPC workloads, increasing amount Big Data/Artificial Intelligence/Machine Learning applications will need to be handled by HPC technology solutions.

Also, the Internet of Things (IoT) will change completely the landscape of using HPC technology, distributing the intelligence and control between the edge and central systems. There is also a clear demand from some domains to use HPC systems for the control of complex systems such as smart grids. The Cloud delivery model is yet another trend affecting the features of future HPC solutions. Accordingly, the SRA has a dimension to address all these new usages.

We also believe that two other areas, namely support mechanisms for SMEs and start-ups as well as education and training need to remain as priorities to help Europe compete in this complex market. These issues are represented by the fourth dimension.

The model below also reflects the importance of the Extreme scale Demonstrators (EsD) as a mechanism to integrate the work of the entire European HPC technology into working system prototypes.

The ETP4HPC HPC Technology research model - it has four dimensions that address the cohesive development of HPC technology.

The process of writing the SRA

The process of preparing SRA 3 began in March 2017. As a first step, we analysed application requirements by collecting input from the following areas:

  • From the Scientific Applications via the Centres of Excellence in Computing Applications (CoEs) and PRACE’s Scientific Case 2017,
  • From Industrial users,
  • From related projects and initiatives, BDEC1 and HiPEAC2, and the Big Data Value Association (BDVA) representing the Big Data community.

In a second step all ETP4HPC members were invited to participate in technical working groups (WGs), mirroring the categories above and led by selected ETP4HPC member organisations. The task of these working groups was to define the research priorities and milestones in the corresponding domains. This work took place through conference calls within the groups and a workshop involving all working group leaders.

Within this process, a special emphasis has been placed on the concept of the Extreme scale Demonstrators – prototypes of European exascale supercomputers. The second area of focus in this SRA is Big Data and its HPC system requirements.

Focus on the Extreme scale Demonstrators

To facilitate the definition of the scope of these systems, a total of five workshops has been held in 2016 and 2017.

There is an emphasis in this 3rd release of the SRA on the concept of European HPC system prototypes, called Extreme scale Demonstrators (EsDs – which form an overarching dimension of the model above). The EsDs are vehicles to optimise and synergise the effectiveness of the entire HPC H2020 Programme through the integration of isolated R&D outcomes into fully integrated HPC system prototypes; a key step towards establishing European exascale capabilities and solutions. The primary focus of the EsD projects will be establishing proof-points for the readiness, usability and scalability potential of the successful technologies developed in previous European actions. To facilitate the definition of these systems jointly with the entire European HPC ecosystem, we held five workshops with the different stakeholders (such as the FETHPC projects, the CoEs, industrial users, and system integrators) to express their comments on the EsDs and to outline their potential contribution.

Extreme-scale demonstrators

The role of the European HPC ecosystem stakeholders in EsD projects.

Focus on Big Data

Being aware of the ever growing importance of highly data-intensive applications, the second area of focus in this third SRA is Big Data and its system requirements. We engaged in a still ongoing collaboration with the BDVA in order to discuss how to create synergies and how to synchronise the technologies of these the two areas.

1. Big Data and Extreme Computing
2. High Performance and Embedded Architecture and Compilation

Our main activities

Our Public Private Partnership with the EC on HPC

The HPC Public Private Partnership: a strategic activity

The HPC cPPP

At the beginning of 2014 we, ETP4HPC, signed the HPC contractual Public Private Partnership (cPPP) with the European Commission. The overarching objective of this cPPP is to develop the next generation of HPC technologies, applications and systems towards exascale, and achieve excellence in HPC application delivery and use. The cPPP governance now actively involves the Centres of Excellence (CoE) launched in 2015, and an active connection is maintained with PRACE as well.

The cPPP can be summarised as a structured dialogue, where the EC commits to striving to secure R&D funding during the H2020 timeframe, while the so-called private side – represented by ETP4HPC - commits to providing recommendations and R&D priorities, matching EC related R&D funding, and jointly monitoring the progress and impact of the overall programme. This latter part of the effort is of utmost importance, to check the impact and effectiveness of the important funding prioritised for HPC. Proper explanations must be delivered to the policy and decision makers, and the activity and effort of ETP4HPC’s members in R&D&I can be highlighted at the same occasion.

Main achievements of the programme

The private side input to the cPPP included three successive releases of an HPC Multi-Annual Roadmap by ETP4HPC (Strategic Research Agendas, or SRA), drawn up in close interaction with the whole EU HPC ecosystem. These SRAs related to the three H2020 Work Programmes (2014-2015, 2016-2017 and 2018-2020). Between 2015 and 2016 FETHPC and CoE projects already received 176 M€ of funding, for a portfolio of 30 projects in hardware and software technologies and various application areas (9 CoE). More projects will start in 2018, selected from 2017 calls. Even more calls have been published for Work Programme 2018-2020: projects selected during this last period of H2020 will actually receive the largest percentage of the cPPP funding.

Scope of HPC cPPP

The cPPP governance actively involves ETP4HPC as well as the Centres of Excellence (CoE)launched in 2015.

H2020 project statistics

2014- 2018 (Q1)
# of H2020 calls implemented 4
Avg. time-to-grant 7 months
Total H2020 funding committed €219.5 million
# of running projects 32
# of new projects to start during the year 2018 11 (RIA) + 2 (CSA)
Projects coordinated by ETP members 19
Participating organisations 429
Unique participations 221
non-ETP members participations 62%
Industry (non-SME) participations Not processed yet
SME participations Not processed yet

Projects funded between 2015 and 2017 under 3 FETHPC and 1 CoE calls – scope of Work Programmes 2014-2015 and beginning of Work Programme 2018-2020

Progress monitoring: KPIs, methodology, data sources

ETPHPC is also required to generate and gather data and create the necessary analysis tools to deliver proper and periodic progress reports; the EXDCI coordination and support action provides extra support for this activity1. There is an agreed list of Key Performance Indicators (KPIs): some are generic and common to all cPPPs and some are specific to HPC. The KPIs fall into three categories:

  • Industrial Competitiveness and Socio-Economy Impact (market shares, jobs, extra private investments…)
  • Operational aspects of the programme (H2020 programme stats but also patents, training activities, software ecosystem etc.)
  • Management aspects of the programme (efficiency, transparency, dissemination etc.)

In order to document the KPIs and assess the impact of the Research & Innovation activities linked to the HPC cPPP strategy, we combined a set of methodologies (covering data gathering, impact and performance metrics, and analysis tools). We involved the stakeholders: PRACE, and ETP4HPC members via an annual survey, FETHPC and CoE projects; and of course the EC is providing a considerable amount of the “operational” data and statistics related to the H2020 project implementation.

We also outsourced efforts for market and complementary professional studies. In particular, collecting socio-economic and business related data from private companies is quite complex and sensitive, but a very necessary task since innovation and value injected into market segments and shares (and jobs) will be one of the most scrutinized aspects of the cPPP final results.

2017 mid-term review - main findings and next steps

In 2017 the EC organised a mid-term assessment of the cPPPs. Our 2014 to 2016 progress reports were of course key input in this process, as was various other activity documentation. The general outcome of the global review was positive: cPPPs are considered relevant, useful instruments, but more time will be needed to assess the socio-economic impact fully.

Regarding the cPPP HPC in particular, our progress reports acknowledged significant ecosystem development, increased industrial participation (in particular compared with FP7) and a good portfolio of projects started in 2015, then extended in 2016. Promising, but still early effects have also been observed in job creation, patents, and innovation from large and small companies. More efforts and progress are expected from the role and involvement of SMEs in the ecosystem and the H2020 projects in general. Finally, the largest fraction of H2020 funding still has to be injected under Work Programme 2018-2020, the actual effects of which will be measurable later on.

One result of this mid-term evaluation is revised KPIs: the EC has started a process with all cPPPs for this purpose, streamlining it to a reduced set of meaningful socio-economic indicators. This is work in progress and should lead to a new organization of progress reports from 2018 onwards.

1. Work Package 7 of EXDCI mobilises a team with ETP4HPC Office, Scapos, and Fraunhofer ITWM participants

Our main activities

EXDCI: coordination for the European HPC ecosystem

ETP4HPC and many of its members contribute to the European H2020 funded project EXDCI, which aims at coordinating the efforts of the European HPC ecosystem.

EXDCI Technical Meeting in Bologna (Italy) in July 2017

The goal of the coordination and support action EXDCI (European eXtreme Data and Computing Initiative) is to “coordinate the development and implementation of a common strategy for the European HPC Ecosystem in order to achieve its global competitiveness within the Horizon 2020 Programme”. ETP4HPC and PRACE are the two main consortium partners of the project. Both organisations bring in a significant number of their partners to support the aims of the project. Today, the European ecosystem gathers hundreds of organisms, research labs, universities, SMEs and larger companies. These entities are involved in numerous European projects (e.g., FETHPC projects, other ICT related projects, or Centers of Excellence) and initiatives, such as BDVA (Big Data Value Association), which should all contribute to the success of the European Cloud Initiative. This multitude of actors and initiatives requires development of a common European HPC Strategy and operation of a synchronized European HPC Community. EXDCI is also facing the data deluge that has created a new focus so that new capabilities for doing science can be developed (e.g., machine learning) and connections to the Big Data market can be made.

The EXDCI approach to Ecosystem coordination

EXDCI addresses the Ecosystem coordination using two approaches. The first method is hierarchical via PRACE and the EPT4HPC while the second is transversal.

The hierarchical method aims at producing roadmaps and identifying the scientific challenges to be addressed: ETP4HPC’s Strategic Research Agenda (SRA) on the one hand and PRACE’s Scientific Case, on the other hand. The Extreme scale Demonstrator proposal combines these two main aspects.

The transversal approach in EXDCI was implemented by a series of actions. The first major initiative by EXDCI was the European HPC Summit Week that was held in Prague (2016) and Barcelona (2017). These events drew hundreds of academics from FETHPC projects, from the Centers of Excellence (CoE), and businessmen. The 2018 edition will take place in Ljubljana (Slovenia) in May.

The second major action was its representation in the international Big Data and Extreme Computing initiative (BDEC). BDEC is an international initiative, with workshops to allow scientists to exchange on roadmapping efforts in order to understand the paradigm shift related to extreme data and computing issues. Previous efforts that started with the IESP (International Exascale Software Project) initiative have been very successful in helping the exascale roadmap definition. EXDCI has participated in the organization of a set of Birds-of-a-Feather sessions and workshops at major events such as Supercomputing (USA) and ISC (Germany).

Moreover, EXDCI took the opportunity to build strong links with the BDVA via technical workshops.

EXDCI has also been working on how to help SMEs and start-ups. In a first phase, we focused on how start-ups and SMEs in HPC perceive their situation and what obstacles they face. In a second phase, we broadened the scope by integrating the point of view of the other stakeholders. This led to a set of recommendations with the goal of strengthening the partnerships between SMEs, start-ups, large companies and HPC centers.

Other transversal actions include monitoring the ecosystem to capture its evolution, synchronizing with Eurolab4HPC, and promoting HPC by publishing Career Case Studies. The purpose of the latter is to promote the various interesting and exciting career opportunities that HPC can offer young people.

Ecosystem Level Recommendations

Better research instruments - The following recommendations aim to improve current research instruments, both computing resources and deployment of new technologies in order better to support applications and researcher discovery processes.
R1: Design new operation policies and federation towards convergence.
R2: Reinforce Big Data and extreme scale international initiatives.
R3: Improve access to advanced technologies.

R&D efficiency - The purpose of the following recommendations is to ensure that public and private investment in R&D is carried out coherently, maximizing thus the impact of research
R4: EuroHPC1 for advanced research and innovation.
R5: Paving the way from Extreme scale Demonstrators (EsD) development to applications.
R6: Improving result capitalization of research projects, such as FETHPC and Centres of Excellence (CoE).

Industry competitiveness - The ambition of the recommendations is to leverage R&D excellence and translate its output into industry competiveness:
R7: Encouraging commercial relationships between SMEs and industry via European projects
R8: Developing a concerted approach to HPC training in Europe
R9: Creating incentives to increase EU stakeholders’ implications in standard initiatives

These recommendations complement the technical ones proposed in the Strategic Research Agenda and form the PRACE Scientific cases.

We are facing a “paradigm shift” driven by the data and computing convergence. It carries many opportunities but also requires the European HPC community to adapt and invent new ways for science and industry. EXDCI has contributed to setting new goals and challenges to the HPC community. This large-scale effort is a first step to bridge applications and technology while encouraging the dialog between all the stakeholders.

1. The European High-Performance Computing Joint Undertaking

Our main activities

Making HPC, Big Data & IoT work together

ETP4HPC joined forces with the Big Data Value Association (BDVA) in amplifying value creation for Europe.

Global Ship Tracking Intelligence (www.marinetraffic.com)

Increasingly large-scale industrial and scientific systems require the aligned use of Internet of Things (IoT), Big Data (BD) and High Performance Compute (HPC) technologies to generate real-time and actionable insights. This need requires a more integrated research approach across HPC-IoT-BD platforms.

To this end, a new collaboration has being initiated between the European associations ETP4HPC, BDVA, and AIOTI1 (representing more than 500 European organisations across the 3 associations). The HiPEAC2 project and the BDEC initiative also participate in this collaborative effort – lending their specific domain expertise to the cause. The objective is to enable the creation of a world-class HPC-BD ecosystem based on European leadership in HPC, Cloud, Edge and Big Data technologies.

Of course, the respective communities will continue independently identifying and researching idiosyncratic challenges particular to their respective domains, but not in isolation from each other. There is growing interdependency between the three domains when looking at future use scenarios like “autonomous driving”, “personalised medicine” or “power grid management & optimization”. HPC will soon be playing a greater role in very dynamic workflows captured under the term “HPC in the loop”. The joint work will ultimately result in a set of joint R&I roadmap elements.

During 2017, the collaboration was already busy exploring the major issues of commonality via documentation of industrial and scientific use cases in the area of “extreme compute and extreme data”. To date, 15 use cases have been identified in the wake of two workshops (Bologna and Versailles). These use cases involving the above mentioned three primary communities will help to shape technical research opportunities across the various compute & data platforms. The results will be the basis for defining new research recommendations offered to the European Commission, such as a proposal for a second EsD call, as well as input to the upcoming 9th Framework Programme.

1. Alliance for Internet of Things Innovation
2. High Performance and Embedded Architecture and Compilation

Our main activities

BDEC Report on the Convergence of Big Data and HPC

According to the just-released report of the BDEC consortium , the Big Data revolution is about to transform our conception of future HPC and Cloud platforms radically.

Global Ship Tracking Intelligence (www.marinetraffic.com)

Big Data and Extreme Computing (BDEC) is an international consortium that, since 2013, has been organising annual workshops, the aim of which is systematically to map out and describe how Big Data-related issues intersect with, impinge upon, and may potentially change, currently envisaged national (and international) plans for achieving exascale computing. Numerous ETP4HPC members have actively participated in drafting the white papers and final report on BDEC.

Bridging the gaps between HPC and Big Data

The primary motivation for the BDEC initiative was the widespread acknowledgement of a growing split between the software ecosystems of traditional HPC and emerging high-end data analysis (HDA), and machine learning technologies. The latter are based largely on open-source products from commercial cloud vendors and are designed to run on those infrastructures. At a time when scientific communities are striving to become more international, more interdisciplinary, and more collaborative, they are facing a host of conceptual, political, economic, and cultural challenges arising from these differences.

But as the BDEC workshops progressed, an even more serious problem became apparent. The landscape of scientific computing is being radically reshaped by the explosive growth in the number and power of digital data generators, ranging from major scientific instruments to the Internet of Things (IoT), and by the unprecedented volume and diversity of the data generated. The most rapidly expanding front of this data tsunami is occurring in “edge environments” (i.e., across the network from both HPC and commercial cloud machine rooms), where the processing and buffer/storage resources for managing such massive data flows are almost entirely lacking. This enters clearly into ETP4HPC scope and its relations with theBig Data Value Association (BDVA).

Pathways to Convergence

In “Pathways to Convergence: Towards a Shaping Strategy for a Future Software and Data Ecosystem for Scientific Inquiry”, the BDEC community report summarizes the exchanges over the last 4 years and examines the progress toward, or potential for, convergence at three different levels.

Examples of successful convergence in large scale scientific applications: several different application communities have already shown that HPC and HDA technologies can be combined and integrated into multi-stage application workflows in order to open new avenues of research.

Progress toward convergence on large scale HPC platforms: The potential for merging simulation-centric and data analysis–centric methods has led to the substantially more flexible forms of system management that such workflows will require.

Absence of community convergence on a next generation distributed services platform: The erosion of the classic Internet paradigm due to Big Data – and the proliferation of competing approaches to create a new type of distributed services platform (e.g., “edge” or “fog computing”) to replace it – is perhaps the most important challenge that the cyberinfrastructure community will confront in the coming decade. The BDEC report is available at: www.exascale.org/bdec/report

Our main events

Major events

7th General Assembly of ETP4HPC @ IBM, Munich

On 21st March 2017, we held our General Assembly in Munich. IBM hosted this event in its new Watson IoT Center. The first focus was on our activities. To this end, our chairman Jean-Pierre Panziera presented an overview of ETP4HPc’s past and future activities, as well as the upcoming evolution of the European HPC landscape. Our Work Groups then reported on their activities. Talks by invited guests complemented this part: Gustav Kalbe (DGCNECT) presented and discussed the EC’s strategy on HPC with ETP4HPC’s members, and Jean-François Lavignon reported on the activities of the international initiative BDEC.

Our members and our ecosystem were our second focus: The General Assembly was also a good opportunity for our new members to introduce themselves, Will Schilders (TU Eindhoven) gave a short presentation on the EU-MATHS-IN initative and possible collaborations with ETP4HPC, and Gordon Wollgam (Roland Berger Consulting) presented a currently running HPC Business case study on behalf of the European Investment Bank (EIB).

Moreover, we began the work on the third edition of our Strategic Research Agenda (SRA3), as guiding reference for the HPC-related calls in the 2018-2020 Work Programme.

European HPC Summit Week in Barcelona

EXDCI coordinates the conference series “European HPC Summit Week”. Its aim is to gather all related European HPC stakeholders (institutions, service providers, users, communities, suppliers and consultants) in a single week to foster synergies. Each year, EXDCI opens a call for contributions to all HPC-related players who would like to participate in the week through a workshop. After Prague in 2016, the 2nd edition of the European HPC Summit Week (EHPCSW17) took place from May 15th to 19th in Barcelona on the Campus of the Universitat Politècnica de Catalunya. The event hosted the EXDIC workshop, as well as the fourth edition of the PRACE Scientific and Industrial Conference on “HPC for Innovation: when Science meets Industry” with international keynotes, panel discussions, and many parallel sessions. In addition, EuroLab4HPC held a workshop, as did ETP4HPC with a session on applications for Extreme scale Demonstrators. The workshops organised by the Centres of Excellence (CoE) and the FETHPC projects ranged from application areas including renewable energies, oil & gas, biomedicine, big data, mathematics, climate modelling, computing applications, to HPC future technologies.

ISC17 and Forum Teratec 2017

The 2017 edition of the ISC High Performance (ISC17) was held in Frankfurt (Germany) from June 19th untill 22nd 2017. Like in previous years, ETP4HPC took part in different ways: First, we presented our SRA and other activities on our booth and discussed with members and visitors the development of the European HPC ecosystem. Moreover, we hosted a workshop focusing on industrial usage (and users) of the future Extreme scale Demonstrators (EsDs) on Thursday (June 22nd 2017).

Another important event in June is Forum Teratec: On June 27th and 28th, Ecole Polytechnique in Palaiseau (France) hosted the Forum Teratec 2017. This forum is one of the major events in France and Europe, which is gathering top international experts in High performance computing, Big Data and simulation. ETP4HPC hosted on its booth EXDCI, and jointly we promoted and explained to visitors the European efforts in HPC.

ETP4HPC and BDVA Joint Workshops

ETP4HPC co-organised two technical exchange workshops with Big Data Value Association (BDVA) with the view of improving a common technical language between their HPC and Big Data platforms and also clarifying the perspective on common subjects such as compute infrastructure and requirements for its next generation.

The main focus of the first workshop in Bologna (July 4th 2017) was on developing a better understanding of the intersections in scope, research directions and priorities for both platforms, with a central discussion on developing a commonly accepted view on the differences between typical HPC compute systems and that of Big Data. Moreover, the participants agreed to identify use cases of common interest, i.e., use cases using both HPC and Big Data technologies. The 2nd workshop brought in representatives from BDEC, HiPEAC and AIOTI – along with BDVA and ETP4HPC. The objective of this workshop was to verify the use cases of common interest and also begin a discussion on preparations for future joint research priorities and roadmaps. The workshop took place on November 20th 2017 in Versailles, organised as part of the European Big Data Value Forum 2017.

EXDCI Final Conference in September 2017 in Barcelona

EXDCI (European eXtreme Data and Computing Initiative) co-located its Final Conference (September 7th - 8th 2017 in Barcelona, Spain) with the Annual ACM Europe Conference 2017, gathering thus the ACM Europe community, HPC specialists, students and ACM members worldwide. Sergi Girona (BSC) as Project Coordinator and François Bodin (Université Rennes I) as EXDCI Technical Director opened the conference with an overview of EXDCI’s activities and achievements. Marcin Ostasz (BSC) and Stéphane Requena (GENCI) reported on upcoming evolutions regarding technologies and applications, respectively; The global trends in HPC and the international competition were sketched by Mark Asch (Université de Picardie), and David Henty (EPCC) presented EXDCI’s efforts on Training for the next generation of HPC experts.; Maike Gilliot (ETP4HPC/Teratec) gave some insight on challenges the SMEs in HPC face and Jean-Philippe Nominé (ETP4HPC/CEA) showed the evolution of the European HPC ecosystem over the last few years. The conference also included a joint session with Eurolab4HPC (H2020 funded coordination and support action) and HiPEAC, represented by Keon De Bosschen, as well as ACM’s track on HPC and a Turing Lecture.

Birds-of-a-Feather Session @ SC17 (Nov. 14th 2018, Denver, Colorado)

For the third time, we held a Birds-of-a-Feather session (BoF) at the International Conference for High Performance Computing, Networking, Storage and Analysis (SC). This years’ edition was entitled “European Exascale Projects and Their Global Contributions”. It was organised by ETPHPC on behalf of the European eXtreme Data and Computing Initiative (EXDCI). The objective of the BoF was to present an overview of the European HPC landscape and the international potential of the European HPC projects, many of which were present at the event with the aim of starting new collaborations.

After a short introduction by Jean-Pierre Panziera, Chairman of ETP4HPC, Eugene Griffiths, (BSC and member of ETP4HPC’s Steering Board) presented an overview of “The European HPC Strategy”. This was followed by presentations of three selected European HPC Projects: BioExcel (Erwin Laure, KTH), SAGE (Sai Narasimhamurthy, Seagate) and Mont-Blanc (Etienne Walter, Atos).

François Bodin (Scientific Director, EXDCI), Paul Messina (Exascale Computing Project, USA) and Yutaka Ishikawa (Riken, Japan) presented an update of the HPC ecosystem in Europe, the US and Japan respectively.

Participation to events

Apart the events and workshops (co-) organised by ETP4HPC, representatives of the association participated actively in different external events in 2017.

Date Event Location Participation of ETP4HP
16-17/01/ 2017 ARM: on the road to HPC, organised by Mont Blanc Project Barcelona, Spain Presentation of ETP4HPC by J.-P. Panziera
30/01/2017 Stakeholder Forum for the Digitising European Industry imitative Essen, Germany ETP4HPC represented by E. Griffiths (BSC) and G. Lonsdale (Scapos)
28/02 - 01/03/ 2017 HPC User Forum HLRS, Stuttgart, Germany M. Malms presenting ETP4HPC and the SRA
08/03/2017 BDEC Workshop Wuxi, China Presentation on "Convergence at European and International levels" by ETP4HPC Chairman J.-P. Panziera
23/03/2017 Official Launch of the European Platform of National Initiatives on Digitising Industry (EUI 4.0) Rome, Italy ETP4HPC Chairman and ETP4HPC Secretary invited
28/03/2017 ORAP Forum Paris, France Presentation of ETP4HPC by J.-P. Nominé
27/04/2017 CompBioMed meeting on Cloud & HPC in Biomedicine Barcelona, Spain Co-located with ACM Europe Conference, UPC
London, UK Hosted by UCL, ETP4HPC represented by H. Falter Brussels, Belguim Hosted at the EC’s office at Beaulieu 33
01-02/06/2017 Innovative Enterprise week: Access to Finance for Research Innovation & SMEs Malta ETP4HPC represented by H. Falter
15-16/06/2017 Digital Assembly 2017: "Digital Europe: Investing in the Future" Valletta, Malta ETP4HPC represented by H. Falter
19-21/06/2017 Booth at the ISC17 Conference Frankfurt, Germany Booth No. 419
27-28/06/2017 Booth at the Forum Teratec Palaiseau, France Booth No. 10
06/07/2017 Exascale State of the Art workshop for the E-CAM Centre of Excellence Barcelona, Spain Presentation of ETP4HPC by M. Ostasz
06/10/2017 Presentation of the NESUS Research Roadmap Brussels, Belgium M. Ostasz representing ETP4HPC
26/10/2017 2nd JPI Oceans Conference Lisbon, Portugal ETP4HPC represented by M. Castrillo (BSC)
9+10/11/2017 ICT Proposer's day Budapest, Hungary Participation of M. Ostasz at the cPPP Discussion
14/11/2017 BoF session at SC17 on European Exascale Projects Denver, USA Organised by ETP4HPC on behalf of EXDCI
15/11/2017 HPC Connects Workshop organised by Asia Supercomputer Community Denver, USA Presentation of the European Exascale Projects by M. Ostasz
21/11/2017 High-level governance meeting of the Platform of National Initiatives on Digitising Industry (led by Commissioner Mariya Gabriel) Bruxelles, Belgium H. Falter representing ETP4HPC
5-7/12/2017 The European Forum for Electronic Components and Systems Brussels, Belgium Representation of ETP4HPC by M. Ostasz

Supporting technology uptake and entrepreneur-ship

SMEs are the backbone of Europe’s economy. To breed and to strengthen start-ups and SME is vital for Europe’s competitiveness. Amongst the multitude of regional, national and European support actions for start-ups and SME, we highlight here some initiatives focussing on HPC and related domains.

EUROLAB4HPC Business Prototyping

EUROLAB4HPC has the bold overall goal to join forces among excellent research institutions towards a long-term research agenda driving innovation and education in HPC. Its program for innovation acceleration aims at validating technical and business hypotheses regarding commercial opportunities for a new technology (Business Prototyping). EUROLAB4HPC has arranged two open calls and mentored six teams. In the future, it will collaborate closely with the TETRAMAX project.

EUROLAB4HPC is a two-year project that aims at promoting excellence in research in HPC systems. One of its key activities is acceleration of innovation in HPC technologies. It has run a program, called Business Prototyping, in which six teams have been mentored. The purpose of Business Prototyping is to help research groups in HPC to identify and evaluate business cases based on their research results. Around each technology, Business Prototyping Project teams are formed that consists of the principal investigator, an entrepreneurial lead, and a mentor with industry experience.

Each team is guided through a method of how to quickly and effectively identify and test critical business-related hypotheses, e.g., who is the customer; what problem are we solving for them; what should our offering be; how and how much should we charge for it. While good technology is critical, Business Prototyping test business related hypotheses, e.g., by talking to customers and other external stakeholders. In many cases, Business Prototyping has made the projects able to attract funding for getting operational in a start-up company.

The next edition of EUROLAB4HPC will promote entrepreneurship through general purpose entrepreneurial training, Business Prototyping, business plan development and helping with funding. Moreover, it will stimulate technology transfer by connecting with other technology transfer activities and providing competitive seed funding for HPC technology transfer.

Fortissimo

Many SMEs understand the business benefits and opportunities that can be identified through digital simulation. Ultimately such simulations can have a huge impact on competitive positioning as well as saving companies time and money. On the other hand, having access to the necessary resources required for digital simulation, both technical & financial, has often been a huge challenge for those SMEs. This is where Fortissimo comes in. The aim of Fortissimo is to offer companies easy, cloud-based access to computational intensive simulations.

Fortissimo, falls within the I4MS initiative (ICT innovation for manufacturing SMEs), which is funded by the European Commission.

Within the project, the Fortissimo Marketplace was launched. The Marketplace provides SMEs with a one-stop-shop for access to hardware, software & expertise specifically for computational intensive simulations. The benefits of offering such solutions means SMEs can decrease time to market, save costs on development, production and operation as well have the capacity and ability to produce more precise and complex designs for components and systems. The one-stop-shop is based upon a pay-per-use set up. The cloud-based Fortissimo Marketplace matches service providers to user’s requirements, handling a range of tasks including project management and billing.

In the past Fortissimo have run “open calls” lasting for a duration of 18 months. Here SMEs proposed potential experiments and those selected to participate are provided by the Fortissimo partners with free access to HPC resources via a cloud infrastructure as well as support with planning, implementation and realization of their experiment.

If you are a manufacturing SME you should check out the Fortissimo Marketplace! Get inspired with our success stories: https://www.fortissimo-project.eu/success-stories

HPC-Europa3 – Transnational visit programme

HPC-Europa3 is a multi-disciplinary programme for short collaborative research visits with transnational access to some of Europe’s biggest supercomputing resources. The programme’s goals are to provide (1) access to world-class HPC systems, (2) technical support by the HPC centres, and (3) scientific collaboration with host researchers in any field. Travel and living expenses will be reimbursed. Over the 4 years of the project, we expect to support more than 1200 visitors and to offer about 100 million CPU hours.

HPC-Europa3 aims to implement a programme of transnational access visits, giving researchers the opportunity to spend a period of time at one of the HPC partners, and to benefit from using some of the most relevant HPC infrastructures in Europe complemented by technical on-site support. Visitors can work closely with a «host» research group in a similar field of research: new «host» can join the programme at any time; SMEs are welcome to participate too. Applicants can be working in any discipline, but must require the use of HPC resources for their project. HPC-Europa3 covers travel and living costs, and visits can last between 3 weeks and 3 months. The programme is open to researchers of all levels, from postgraduate students to senior professors. Applications are encouraged from the new EU member countries and those who do not have access to similar computing facilities at their home institute.

Under the new «Regional Access» sub-programme, applicants in the Baltic states and South East Europe who have little or no HPC experience are particularly encouraged to visit Sweden and Greece sites respectively. Calls are open all the time but evaluations take place 4 times per year. In parallel, HPC-Europa3 Networking Activities promote co-operation with other European initiatives and Joint Research Activities address technological issues to improve the performance of research infrastructures. HPC-Europa3 receives funding by EU H2020 programme under grant agreement No.730897.

SHAPE: Removing barriers to HPC for SMEs

Small and medium-sized enterprises (SMEs) are the backbone of the European economy. Many could take advantage of the opportunities available from adopting High Performance Computing (HPC) in their business, but most do not. SHAPE is a European programme designed to help SMEs overcome barriers to HPC adoption.

SHAPE (SME HPC Adoption Programme in Europe) is a pan-European programme that promotes High Performance Computing (HPC) adoption by SMEs (Small and medium-sized enterprises), supported as part of the PRACE (Partnership for Advanced Computing in Europe) initiative.

It can be challenging for SMEs to adopt HPC. They may have no in-house expertise, no access to hardware, or be unable to commit resources to a potentially risky endeavour. This is where SHAPE comes in, by making it easier for SMEs to make use of HPC in their business - be it to improve product quality, reduce time to delivery or provide innovative new services to their customers.

Successful applicants to the SHAPE programme get effort from a PRACE HPC expert and access to machine time at a PRACE centre. In collaboration with the SME, the PRACE partner helps them try out their ideas for utilising HPC to enhance their business.

So far SHAPE has helped 39 companies to benefit from using HPC in their business (see the project website for examples), and with the recently closed 6th call under review, many more will get assistance from PRACE. The companies cover a diverse range of areas, including prosthetics, electromagnetic simulation, gas burners and propeller design.

The next SHAPE call opens on 1st April 2018 and there will be further calls during the PRACE project, so there are plenty of opportunities for SMEs to take part and try out HPC in their business.

TETRAMAX

The mission of TETRAMAX is to establish a European competence center network in the domain of customized and low-energy computing (CLEC) for cyber-physical systems and the Internet of Things as well as support cross-border Technology Transfer Experiments through a variety of open calls. In the long term, TETRAMAX will be the trailblazer towards a reinforced and sustainable ecosystem infrastructure, providing CLEC competences, services and a continuous innovation stream at European scale, yet with strong regional presence as preferred by SMEs.

To become or stay globally competitive, European companies need immediate and low-risk access to the latest embedded computing technologies. This green and low-energy computing creates opportunities in digitalization which will lead to improved products and higher productivity. As a new Horizon 2020 innovation action, TETRAMAX will provide an implementation of the European Smart Anything Everywhere initiative, thereby building on three major activity lines: (1) stimulating, co-funding, and evaluating different cross-border Technology Transfer Experiments (TTX) via innovative CLEC technologies to first-time users and broad markets in European ICT industries, (2) building and leveraging a new European CLEC competence center network, offering technology brokerage, one-stop shop assistance and CLEC training to SMEs and mid-caps, with a clear path towards new regional Digital Innovation Hubs, and (3) paving the way towards self-sustainability.

Within the framework of different open calls for TTX proposals, the immediate ambition is to support 50+ industry clients all over Europe with innovative technologies, leading to an estimated revenue increase of €25m based on 50+ new/improved CLEC-based products, 10+ entirely new businesses/SMEs initiated, as well as 30+ new permanent jobs and significant cost savings in product manufacturing.

In November 2017 TETRAMAX published its 1st call for bilateral Technology Transfer Experiments.

Call details: www.tetramax.eu/ttx/calls

Credits

Text
ETP4HPC
ETP4HPC Chairman
Jean-Pierre Panziera
Editorial Team
Maike Gilliot (Coordinator) - Michael Malms Jean-Philippe Nominé - Marcin Ostasz
External contributions
  • Support actions for SMEs and start-ups: HPC-Europa3, EuroLab4HPC, Fortissimo, PRACE, TETRAMAX
  • New members: 2CRSI, Asperitas, ESI Group, IFP Energie Nouvelles, ICEOTOPE, Heidelberg University Computing Centre, HPCNow!, NEC, ProDesgin, University of Zagreb, Verne Global
  • Article on BDEC: Mark Asch
  • Article on EXDCI: Francois Bodin
  • Article on BDVA: Jim Kenneally
  • Editorial: Gustav Kalbe
Graphic design and layout
Antoine Maiffret (www.maiffret.net)
Web development
Laurent Mercier
This site requires Javascript to work, please enable Javascript in your browser