Go to the content (press return)

High performance computing VII

Total activity: 41
Type of activity
Competitive project
Funding entity
Funding entity code
200.355,67 €
Start date
End date
Big Data, BigData, co-diseño tecnologías hw/sw, hw/sw technologies co-design, performance and energy efficiency, rendimiento y eficiencia energética, supercomputación, supercomputing
Besides the necessity to run large numerical simulations, the exponential growth in the amount of available data brings a new paradigm to
the scientific method, which is based on massive data analysis; Big Data refers to the set of technological challenges associated to the
massive collection, management, processing, analysis and visualization of data. These new data-intensive workloads represent a groundbraking
shift that is transforming the entire HPC research landscape, in an era with rapid and disruptive changes in the design of
processor, memory, interconnect and storage technologies. This HPC/BigData convergence requires a revolutionary approach, based on
a strong hardware and software co-design strategy at its centre, for the design of the hardware technologies and software stack, being
energy-efficiency and resiliency clear concerns that need to be addressed.
The first objective of this research team is to influence the way HPC systems are built, programmed and used. To achieve this mission,
this project proposal will 1) explore novel technologies and architectures that can reduce the energy consumption, making use of
heterogeneous core/accelerator designs and hybrid memory organizations to support the the integration of Big Data analytics and
cognitive computing into scientific workflows; 2) explore the combination of high-speed networks and high-performance and high-capacity
memories to accelerate the processing of vast amounts of data by having data closer to processing elements; 3) evolve programming
models to address programmability, portability and easy access to data, at a scale of hundreds of thousands nodes, each with hundreds of
cores; 4) evolve performance analysis tools to extract from huge volumes of data generated from instrumented runs of programs, which
make performance understanding a Big Data problem itself, requiring the use of Big Data infrastructures to store performance data in a
distributed way and to be able to query and analyse it in parallel; 5) execution environments on top of virtualized infrastructures able to
exploit the parallelism out of that concurrency expressed, access to data in distributed storage technologies and optimize energy
utilization; and 6) explore new algorithmic issues to efficiently perform data analytics and in-situ large scale data-driven simulations.
The second objective of this research team is to publish results out of research in the best conferences and journals, building prototypes to
materialize these ideas that can be offered open-source and used on real benchmarks and applications. Cooperation with manufacturers
through research agreements and involvement in European projects are the mechanisms that the research group will continue to use to
exploit the ideas in real products and scenarios.
The trajectory and solid background of the research team, with a critical mass of experienced researchers, spirit to open new exploratory
paths and holistic/vertical background, position itself as a key international leader in the HPC/BigData convergence and application to nonstandard
domains (for example cognitive computing or IoT), with a focus on the technologies that will provide the energy-efficiency
required by future compute- and data-intensive workloads, following the technologies from the rapidly changing HPC and embedded
Adm. Estat
Plan Estatal de Investigación Científica y Técnica y de Innovación 2013-2016
Call year
Funcding program
Programa Estatal de Fomento de la Investigación Científica y Técnica de Excelencia
Funding subprogram
Subprograma Estatal de Generación de Conocimiento
Funding call
Excelencia: Proyectos I+D
Grant institution
Gobierno De España. Ministerio De Economía Y Competitividad, Mineco


Scientific and technological production

1 to 41 of 41 results