Grant: Discovery Grants Program - Individual
University of Victoria
$33,000.00
Multicore processing: Structured algorithm parallelization techniques & optimized network-on-chip design to connect to cores
EF 2011-12
Résumé
Multicore processors are used to improve microprocessor performance while limiting power consumption. This technique is used in cell phones all the way up to server farms. Multicore design is able to satisfy these requirements and this justifies the intense research interest in this computer architecture paradigm. This new computing model gives rise to two very important challenges:
(1) ) Software programs and applications must now be rewritten in a parallel programming model to keep all the cores busy and attain the hoped-for performance. As a result, application developers and programmers must learn how to write parallel code. Surprisingly, this effort is still ad hoc and heuristic.
(2) ) The multicores need to communicate among themselves using an efficient and fast interconnection network that replaces the standard system bus with a network-on-chip (NoC).
This proposed research focuses on the above two challenges as the two main areas of research activity:
(1) Mapping parallel algorithms on multicore chips through multithreading. We are developing structured nonlinear techniques to explore the inherent parallelism in different algorithm classes. Programmers will have a powerful tool to program multicores processors. Our structured techniques are necessary to enable programmers to explore options for parallel algorithms for multicores.
(2) Designing efficient Network-on-Chip to connect the different processing cores on the multicore chip. This allows the separation between processor design and communication requirements. This research will allow designers to use novel NoC architectures where several NoC are used simultaneously. Furthermore, we modify the traffic patterns of the modules to accommodate the NoC architecture being used. We use theoretical modeling and numerical multi-objective optimization techniques to study the design space of our NoC. Several optimization techniques have been tried for comparison of results.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
British Columbia, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
4488-2011
Name nom
Gebali, Fayez
Department département
Electrical and Computer Engineering
Organization id
7
Institution établissement
University of Victoria
Province en
British Columbia
Province fr
Colombie-Britannique
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2011
Competition year année de concours
2011
Award amount
33000
Program id
RGPIN
Program nane en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
Discovery grants programs
Group fr
Programmes de subventions à la découverte
Committee code
1507
Committee name en
Computer Science
Committee name fr
Sciences informatiques
Area of application code
803
Area of application group en
Information and communication services
Area of application group fr
Services d'information et de télécommunications
Area of application en
Computer communications
Area of application fr
Téléinformatique
Research subject code
2508
Research subject group en
Electrical and electronic engineering
Research subject group fr
Génie électrique et électronique
Research subject en
Communications networks
Research subject fr
Réseaux de télécommunications
Partie
1
Nb partie
5
Application title
Multicore processing: Structured algorithm parallelization techniques & optimized network-on-chip design to connect to cores
Keyword
NSERC
Application summary
Multicore processors are used to improve microprocessor performance while limiting power consumption. This technique is used in cell phones all the way up to server farms. Multicore design is able to satisfy these requirements and this justifies the intense research interest in this computer architecture paradigm. This new computing model gives rise to two very important challenges:
(1) ) Software programs and applications must now be rewritten in a parallel programming model to keep all the cores busy and attain the hoped-for performance. As a result, application developers and programmers must learn how to write parallel code. Surprisingly, this effort is still ad hoc and heuristic.
(2) ) The multicores need to communicate among themselves using an efficient and fast interconnection network that replaces the standard system bus with a network-on-chip (NoC).
This proposed research focuses on the above two challenges as the two main areas of research activity:
(1) Mapping parallel algorithms on multicore chips through multithreading. We are developing structured nonlinear techniques to explore the inherent parallelism in different algorithm classes. Programmers will have a powerful tool to program multicores processors. Our structured techniques are necessary to enable programmers to explore options for parallel algorithms for multicores.
(2) Designing efficient Network-on-Chip to connect the different processing cores on the multicore chip. This allows the separation between processor design and communication requirements. This research will allow designers to use novel NoC architectures where several NoC are used simultaneously. Furthermore, we modify the traffic patterns of the modules to accommodate the NoC architecture being used. We use theoretical modeling and numerical multi-objective optimization techniques to study the design space of our NoC. Several optimization techniques have been tried for comparison of results.