Numerical simulation and modelling using High-Performance Computing (HPC) evolved into a key research methodology in both academia and industry and the need to solve problems of ever-growing complexity intensifies the quest to attain capable exascale computing. The challenges...
Numerical simulation and modelling using High-Performance Computing (HPC) evolved into a key research methodology in both academia and industry and the need to solve problems of ever-growing complexity intensifies the quest to attain capable exascale computing. The challenges brought about this quest urgently need to be addressed in the coming years, in order to maintain the excellence of EU in research and innovation.
HPC-LEAP implements an interdisciplinary strategy for educating the next generation of scientists to address these challenges, by bringing together world-leading experts to train researchers in Europe in mathematical and computational concepts underpinning current and future simulations in turbulent flows, computational biology and lattice quantum chromodynamics (QCD).
The research projects are designed to enhance collaborations and interactions across these disciplines, integrate non-academic partners and to develop methodologies that efficiently use large-scale simulations on future HPC systems. The main objectives of the programme are to:
- Provide a rigorous research programme, which is interdisciplinary and at the same time has the required depth, by engaging world-leaders in computational science domains in a collaborative environment.
- Contribute to developing a community of scientists in Europe that use simulation to bridge experimentation and theory, by designing innovative HPC-driven research projects and training courses.
- Advance the computational capabilities of the different communities by accommodating future HPC architectures requiring higher levels of parallelism and educate scientists to understand the challenges involved in the co-design approach and in managing big data associated with HPC.
- Pioneer innovative HPC-related industries by integrating non-academic partners in the research projects, effectively training young scientists in innovation practices.
The programme will award double degrees agreed between seven institutions in four EU countries while six additional academic and four non-academic partners complement the training.
The project formally began in April 2015, although the project website was in operation since October 2014, publicising the programme and its aims. By October 2015, the fifteen Fellowship positions were filled by highly-qualified candidates, the EJD framework was in place and each Fellow formulated a detailed career development plan to ensure the smooth and timely implementation of the programme. Around the same time, the Fellows started receiving their formal training.
To date, HPC-LEAP organised four core three-week workshops offering a well-balanced mixture of general knowledge on HPC architectures and algorithmic design. These were intensive, research-oriented workshops aimed to equip the Fellows with the skills necessary to tackle the research problems in their thematic areas. Video recordings from the core workshops are hosted on the supercomputing portal of the Cyprus Institute (http://supercomputing.cyi.ac.cy/)
Moreover, three elective, weeklong workshops on computational biology, turbulent flows and lattice QCD took place. All workshops organised within HPC-LEAP were also open to attend for researchers outside the network. Further training is provided depending on the specific needs of the Fellows and following a consultation with their academic supervisors. The Fellows have also broadened their research experience, by liaising with leading scientists through secondments to partner academic institutions and industrial partners, by receiving further training and having access to state-of-the-art technologies and methodologies.
Concurrently with their training, the Fellows engaged in research in the thematic areas of the project. In terms of research dissemination activities to date, most of the Fellows have already presented their work in conferences (poster presentations and/or research talks) and their results have started finding their way into the scientific literature (six peer-reviewed articles published or accepted for publication; six appeared in conference proceedings and six submitted or in preparation). It is anticipated that the efforts to disseminate research outputs will intensify later on, in the more advanced stages of the project.
The Fellows have also participated in a number of outreach activities advocating HPC and advertising the activities of HPC-LEAP to the general public. This includes representing HPC-LEAP at the European Researchers\' Night in 2016, as well as participating in University Open Days. To date, focus has been placed in the training component and research so that the Fellows get started in their projects. More activities are currently planned for the next reporting period.
The potential impact of the programme arises in producing the human capital needed for addressing global issues, such as personalized medicine and safer aircraft, as well as furthering our theoretical understanding of turbulence and unravelling the structure of matter providing insights into the early universe. Undoubtedly, this will also positively affect the competitiveness of European science and engineering research capabilities, as well as enhance the ability of EU industry to pioneer HPC developments towards exascale computing.
More specifically, the training and research components of the programme aim to produce graduates who are:
- Versatile with the mathematical concepts underpinning current and future parallel numerical simulations;
- Skilled software developers taking advantage of large-scale computing resources to advance their research;
- Knowledgeable about the best data analysis techniques to achieve their research aims;
- Confident in presenting their research findings;
- Flexible researchers, able to work on new projects in other disciplines, with a clear overview of contemporary academic and industrial research that uses large-scale simulation;
- Excellent in their chosen research discipline.
- Well-equipped to embark on successful careers in both academia and industry.
Although the research undertaken by the Fellows is still at its early stages, some research highlights over the reporting period include
- The development of a new massively parallel quantum mechanics/molecular mechanics interface, which introduces a paradigm shift than can be utilised for speeding up computations in biomolecular research.
- Gaining insights into highly binding affinity between cisplatin (a potent cancer drug) and DNA using tools from computational biology.
- The development of a novel approach for asynchronous solvers for linear systems, which is an emerging class of highly-scalable, fault-tolerant methods, with potential applications in the exascale regime.
- The development of a new lattice Boltzmann algorithm for the study of problems in the field of relativistic hydrodynamics, which can find applications both at both large (e.g. astrophysics) and small scales (e.g. in quark gluon plasmas).
- The implementation of a multigrid solver for lattice QCD for studying a wide range of observables. It has been used both for the production of ensembles and the computation of physical quantities, giving a significant speedup compared to previous methods.
More info: http://www.hpc-leap.eu.