1. Smart Multicore Embedded Systems SMECY
Francois Pacull (CEA, France)

2. What is SMECY about ? Means To gather European competencies
To provide compilation chains for multi-core platforms
Goal
ASP5 Computing Environments for Embedded Systems

3. SMECY means Project duration: 36 months Start date: 01/02/2010
France
Commissariat à l'énergie Atomique CEA
Thomson Grass Valley France SA GVF
HPC Project HPC
SKYLAB Industries SKYLAB
STMicroelectronics (Grenoble 2) SAS ST-GNB2
Thales Research & Technology TRT-FR
Université Joseph Fourier grenoble 1 UJF/Verimag
Netherland
Associated Compiler Experts bv ACE
Philips Medical Systems PMS
Technische Universiteit Delft TUDelft
Greece
Aristotle University of Thessaloniki AUTH
Hellenic Aerospace Industry S.A. HAI
University of Ioannina UOI
Rep. Czech
Brno University of Technology BUT
CIP plus s.r.o. CIP
Ustav Teorie Informace A Automatizace- Akademie Ved Ceske Republiky Verejna Vyzkumna Instituce Utia
Denmark
Danmarks Tekniske Universitet DTU
Sweden
Free2Move AB F2M
Hogskolan i Halmstad HH
Realtime Embedded AB RTE
Saab Microwave Systems SMW
Finland
Nethawk Oyj NetHawk
Tellabs Oy Tellabs
Valtion teknillinen tutkimuskeskus VTT
Italy
Politecnico di Milano Dipartimento di Elettronica e Informazione POLIMI
Politecnico di Torino POLITO
SELEX Sistemi Integrati SELEX-SI
STMicroelectronics S.r.l. ST Italy Alma Mater Studiorum - Universita di Bologna UNIBO
United Kingdom
Thales Research & Technology TRT-UK
Project duration: 36 months
Start date: 01/02/2010
Total Costs: 20.5 M€
Coordinator: Francois Pacull (CEA)

4. Needs to deal with Different platforms Different application domains
SMECY Goal
Needs to deal with Different platforms Different application domains
Compilation chain depends on
- a platform a class of applications (i.e. type of parallelization) 4

5. 3 clusters Cluster 1 Cluster 2 Cluster 3
Radar signal processing and earth observation
ST P2012
Cluster 2
Multimedia, mobile and wireless transmission
Cluster 3
Stream processing (Video surveillance)
UTIA EdkDSP

6. Place of SMECY in ARTEMIS strategy ?

7. Paradigm changes Multi-core no longer manageable by hand
Would be too complex
Cannot optimise the efficiency
Variability of cores (computing units)
Correction by SW
Way to express / extract / exploit parallelism
Depends on application and platform

8. Domains covered by SMECY
Industrial
Nomadic Environment
Private Spaces
Public Infrastructure
Reference Designs
And
Architecture
Seamless connectivity
Middleware
System Design Methods and Tools
ACROSS
SMECY
POLLUX
R3-COP
ME3GAS
CHIRON
eSONIA
p.S.H.I.E.L.D
RECOMP
SIMPLE
SMARCOS
iFEST
ASAM

9. Cross domain: How SMECY deal with …
Cross domain at application level
Radar
Signal Processing
Earth Observation
Multimedia
Mobile Node
Wireless Transmission
Stream Processing

10. High level targets
50% of ES deployed around the world will be based on ARTEMIS results and will have been developed within the engineering discipline established by ARTEMIS, encompassing hardware, software and systems design for Embedded Systems.
SMECY has the support of a big European semiconductor company specialized in Embedded Systems

11. High level targets
ARTEMIS will have achieved the cross-domain connectivity and communication capabilities necessary to realise the seamless interoperability between the ‘Ambient Intelligent Environments’ envisaged for the European citizen (at home, travelling - in various modes, at work, in public spaces, ...).
Various application domains related to that are considered in SMECY

12. High level targets
There will be twice as many European SMEs within the aegis of ARTEMIS engaged in the Embedded Systems supply chain, from concept through design and manufacture, delivery and support, as there are today.
SMECY has 5 SMEs in the consortium (1/6)

13. High level targets
There will be an integrated chain of European sourced tools, based on ARTEMIS results, to support development of Embedded Systems from user requirements, through system design, to system-on-chip production.
Just what SMECY is about

14. High level targets
ARTEMIS will have generated at least 5 ‘radical innovations’ of a similar paradigm-breaking nature to the microprocessor, digital signal processing and software radio. As a general indicator of innovation, the number of relevant patents granted per annum to European companies engaged in ARTEMIS will have doubled.
Future will tell … :-)

15. High level targets
By 2016, the European research infrastructure and education system will have developed the capability to support the fast dynamic evolution of the embedded systems industry needs in terms of design skills, based on the ARTEMIS recommendations. Major educational programmes and technology acquisition programmes will be able to deliver new skills in less than 2 years.
Universities in SMECY have put or are going to put Multi-core programming as part of their regular cursus

16. All these points are addressed in SMECY
High level targets
To close the design productivity gap between potential and capability:
Reduction of the system design cost (50%)
Reduction in development cycle (50%)
Complexity increase (100%), effort reduction (20%)
Effort and time required for re-validation and recertification (50%)
Cross-sectoral reusability
All these points are addressed in SMECY

17. Innovation
How does SMECY push innovation forward from fundamental R&D to broad industrial applicable innovation ?
Outcome of R&D will be integrated in an indutrial platform

18. Standardisation, Education and Tool platform
How SMECY address the Standardisation, Education and Tool platform ambition described in the ARTEMIS SRA ?
Opensource and Openstandart is under consideration
Eductation aspect is clearly part of the university strategies
Tools from SMECY go in the direction of an European alternative.

19. Thank you for your attention