Transparent heterogeneous hardware Architecture deployment for eNergy Gain in Operation

Información del proyecto

TANGO

Identificador del acuerdo de subvención: 687584

DOI

10.3030/687584

Proyecto cerrado

Fecha de la firma de la CE 9 Diciembre 2015

Fecha de inicio 1 Enero 2016

Fecha de finalización 31 Diciembre 2018

Financiado con arreglo a

• INDUSTRIAL LEADERSHIP - Leadership in enabling and industrial technologies - Information and Communication Technologies (ICT)

Coste total

€ 3 199 625,00

Aportación de la UE

€ 3 199 625,00

Coordinado por

ATOS SPAIN SA
Spain

CORDIS proporciona enlaces a los documentos públicos y las publicaciones de los proyectos de los programas marco HORIZONTE.

Los enlaces a los documentos y las publicaciones de los proyectos del Séptimo Programa Marco, así como los enlaces a algunos tipos de resultados específicos, como conjuntos de datos y «software», se obtienen dinámicamente de OpenAIRE .

Resultado final

Documents, reports (9)

TANGO Requirements and Architecture Specification Final Version

This deliverable is the final version of the architecture and requirements collecting the necessary updates in order to achieve final iteration goals of holistic self-adaptation, considering the final Use Cases requirements.

TANGO Toolbox – Beta version Scientific Report

This deliverable will include progress beyond the state of the art achieved in the Beta version of the TANGO architecture. These will focus on augmenting software modelling tools, code optimisation and programming model adding the capabilities required in order to enable run-time self-adaptations .

TANGO Use Cases Validation – Alpha version

This deliverable will report the alpha version validation for TANGO Toolbox performed in Tasks 6.1, 6.2 on the eHealth Application and 6.3 for Mini-Apps for HPC. It will detail progress in Use Cases architecture and implementation, as well, as validation and verification of TANGO Toolbox components. In addition, it will provide lessons learned from this first iteration that will be used as inputs for the Beta Version of the TANGO Toolbox.

TANGO Toolbox – Alpha version Scientific Report

This deliverable will collect all scientific findings for TANGO alpha version specifically considering work on generic pattern identification, as well as, programming model and metrics and benchmarks advances.

TANGO Requirements and Architecture Specification Alpha Version

This deliverable will contain the results of the state-of-the art and market analysis for the project first iteration. It will contain initial scenarios descriptions and requirements from the Use cases and technical requirements for Static Energy-Efficiency iteration objectives. First version of the TANGO Architecture will be produced based on these.

TANGO Use Cases Validation – Beta version

Similarly to the D6.1, this deliverable will report the beta version validation for TANGO Toolbox done in tasks 6.1-3. In addition to the progress achieved in the iteration, it will include lessons learned .to be used as inputs for the Final Version of the TANGO Toolbox.

TANGO Requirements and Architecture Specification Beta Version

This deliverable will provide an update of D1.1 taking into account updates from the Use Cases requirements, state-of-the-art updates in order to produce requirements and architecture for TANGO Toolbox beta version focusing on Dependability, security and increased productivity.

TANGO Toolbox – Final version Scientific Report

This deliverable will report all scientific finding for the final version of the TANGO Toolbox, considering the optimization and simulation factors for all components in the architecture and considering user’s requirements energy efficiency, performance, security, data location, time criticality, dependability .

TANGO Use Cases Validation –Final version

This deliverable will update previous deliverables produced by the workpackage. In addition to the validation process previously detailed for the other deliverables by the Use Cases applications, the final version will elaborate on lessons learned for project realisation and recommendations for future users.

Other (3)

TANGO Toolbox – Final version

This deliverable will integrate all software components defined in the final architecture of TANGO Toolbox Architecture considering the holistic self-optimisation and across all layers.

TANGO Toolbox – Alpha version

This deliverable will integrate all software components defined in the alpha version of the TANGO Toolbox Architecture considering TANGO iteration 1 ambition of bringing energy awareness to components for heterogeneous parallel architectures in a cross-layer programming flow.

TANGO Toolbox – Beta version

This deliverable will integrate all software components defined in the beta version of the TANGO Toolbox Architecture considering TANGO iteration 2 ambitions of enhancing initial Toolbox version with regards to dependability, security and increased productivity.

Publicaciones

Book chapters (3)

Towards an Energy-Aware Framework for Application Development and Execution in Heterogeneous Parallel Architectures

Autores: Karim Djemame, Richard Kavanagh, Vasilios Kelefouras, Adrià Aguilà, Jorge Ejarque, Rosa M. Badia, David García Pérez, Clara Pezuela, Jean-Christophe Deprez, Lotfi Guedria, Renaud De Landtsheer, Yiannis Georgiou
Publicado en: Hardware Accelerators in Data Centers, 2019, Página(s) 129-148, ISBN 978-3-319-92791-6
Editor: Springer International Publishing
DOI: 10.1007/978-3-319-92792-3_7

Towards Virtual Machine Energy-Aware Cost Prediction in Clouds

Autores: Mohammad Aldossary, Ibrahim Alzamil, Karim Djemame
Publicado en: Economics of Grids, Clouds, Systems, and Services, Edición 10537, 2017, Página(s) 119-131, ISBN 978-3-319-68065-1
Editor: Springer International Publishing
DOI: 10.1007/978-3-319-68066-8_10

Enabling GPU Support for the COMPSs-Mobile Framework

Autores: Francesc Lordan, Rosa M. Badia, Wen-Mei Hwu
Publicado en: Accelerator Programming Using Directives, Edición 10732, 2018, Página(s) 83-102, ISBN 978-3-319-74895-5
Editor: Springer International Publishing
DOI: 10.1007/978-3-319-74896-2_5

Peer reviewed articles (6)

PaaS-IaaS Inter-Layer Adaptation in an Energy-Aware Cloud Environment

Autores: Karim Djemame, Raimon Bosch, Richard Kavanagh, Pol Alvarez, Jorge Ejarque, Jordi Guitart, Lorenzo Blasi
Publicado en: IEEE Transactions on Sustainable Computing, Edición 2/2, 2017, Página(s) 127-139, ISSN 2377-3782
Editor: IEEE
DOI: 10.1109/tsusc.2017.2719159

Towards energy aware cloud computing application construction

Autores: Django Armstrong, Karim Djemame, Richard Kavanagh
Publicado en: Journal of Cloud Computing, Edición 6/1, 2017, ISSN 2192-113X
Editor: Springer Science + Business Media
DOI: 10.1186/s13677-017-0083-2

Energy-aware cost prediction and pricing of virtual machines in cloud computing environments

Autores: Mohammad Aldossary, Karim Djemame, Ibrahim Alzamil, Alexandros Kostopoulos, Antonis Dimakis, Eleni Agiatzidou
Publicado en: Future Generation Computer Systems, Edición 93, 2019, Página(s) 442-459, ISSN 0167-739X
Editor: Elsevier BV
DOI: 10.1016/j.future.2018.10.027

A Holistic Resource Management for Graphics Processing Units in Cloud Computing

Autores: Abdulaziz Alnori, Karim Djemame
Publicado en: Electronic Notes in Theoretical Computer Science, Edición 340, 2018, Página(s) 3-22, ISSN 1571-0661
Editor: Elsevier BV
DOI: 10.1016/j.entcs.2018.09.002

Rapid and accurate energy models through calibration with IPMI and RAPL

Autores: Richard Kavanagh, Karim Djemame
Publicado en: Concurrency and Computation: Practice and Experience, 2018, Página(s) e5124, ISSN 1532-0626
Editor: John Wiley & Sons Inc.
DOI: 10.1002/cpe.5124

Executing linear algebra kernels in heterogeneous distributed infrastructures with PyCOMPSs

Autores: Ramon Amela, Cristian Ramon-Cortes, Jorge Ejarque, Javier Conejero, Rosa M. Badia
Publicado en: Oil & Gas Science and Technology – Revue d’IFP Energies nouvelles, Edición 73, 2018, Página(s) 47, ISSN 1294-4475
Editor: Editions Technip
DOI: 10.2516/ogst/2018047

Conference proceedings (6)

A methodology for efficient code optimizations and memory management

Autores: Vasilios Kelefouras, Karim Djemame
Publicado en: Proceedings of the 15th ACM International Conference on Computing Frontiers - CF '18, 2018, Página(s) 105-112, ISBN 9781-450357616
Editor: ACM Press
DOI: 10.1145/3203217.3203274

Enabling Python to execute efficiently in heterogeneous distributed infrastructures with PyCOMPSs

Autores: Ramon Amela, Cristian Ramon-Cortes, Jorge Ejarque, Javier Conejero, Rosa M. Badia
Publicado en: Proceedings of the 7th Workshop on Python for High-Performance and Scientific Computing - PyHPC'17, 2017, Página(s) 1-10, ISBN 9781-450351249
Editor: ACM Press
DOI: 10.1145/3149869.3149870

Performance and Energy-Based Cost Prediction of Virtual Machines Auto-Scaling in Clouds

Autores: Mohammad Aldossary, Karim Djemame
Publicado en: 2018 44th Euromicro Conference on Software Engineering and Advanced Applications (SEAA), 2018, Página(s) 502-509, ISBN 978-1-5386-7383-6
Editor: IEEE
DOI: 10.1109/seaa.2018.00086

Performance and Energy-based Cost Prediction of Virtual Machines Live Migration in Clouds

Autores: Mohammad Aldossary, Karim Djemame
Publicado en: Proceedings of the 8th International Conference on Cloud Computing and Services Science, 2018, Página(s) 384-391, ISBN 978-989-758-295-0
Editor: SCITEPRESS - Science and Technology Publications
DOI: 10.5220/0006682803840391

Resource boxing - converting realistic cloud task utilization patterns for theoretical scheduling

Autores: Bernhard Primas, Peter Garraghan, Karim Djemame, Natasha Shakhlevich
Publicado en: Proceedings of the 9th International Conference on Utility and Cloud Computing - UCC '16, 2016, Página(s) 138-147, ISBN 9781-450346160
Editor: ACM Press
DOI: 10.1145/2996890.2996897

A unified model for holistic power usage in cloud datacenter servers

Autores: Peter Garraghan, Yaser Al-Anii, Jon Summers, Harvey Thompson, Nik Kapur, Karim Djemame
Publicado en: Proceedings of the 9th International Conference on Utility and Cloud Computing - UCC '16, 2016, Página(s) 11-19, ISBN 9781-450346160
Editor: ACM Press
DOI: 10.1145/2996890.2996896