Grant: Discovery Grants Program - Individual
University of Waterloo
$52,000.00
A Hardware/Software Codesign Approach for Time-Predictable Computing
EF 2024-25
Résumé
Cyber-Physical Systems (CPS) are the next generation of smart, interconnected embedded systems. Many emerging CPS, such as unmanned aerial vehicles and autonomous cars, demand high computing performance for increased autonomy and real-time decision making, while meeting various constraints such as size, weight, power, and cost. CPS are safety-critical systems exhibiting real-time requirements: a failure to compute required actuation commands within a precise time window can lead to catastrophic results. For this reason, computing platforms used in CPS must be predictable, i.e., analyzable to derive tight bounds on worst-case task execution time. Unfortunately, existing multicore computing infrastructures are fundamentally inadequate to meet such timing constraints: their poor predictability - especially due to contention for access to shared hardware resources such as DRAM main memory and caches - is a major hurdle that makes verification and safety certification extremely challenging.
Existing research has tackled such predictability problem by introducing a set of software-based techniques to partition hardware resources between contending tasks or Virtual Machines (VMs). However, such approaches have fundamental limits as they do not directly influence resource allocation and arbitration decisions made at the hardware level. Consequently, predictability is typically achieved at the cost of performance degradation.
In this proposal, I argue that the key to maintaining predictability without performance compromises is a hardware / software codesign approach where hardware controllers expose suitable "control knobs" that can be tuned by the system software (OS / hypervisor) based on applications' requirements. At the hardware level, I will demonstrate the applicability of this concept by modifying both the shared cache and main memory controller, and building suitable control mechanisms in the CPU core. At the software level, I will design and implement suitable interfaces between applications, the OS and the hypervisor. Compared to a general-purpose system, in an embedded setting the system integrator typically has a precise knowledge of the applications' behavior, and such information can be exploited to optimize resource arbitration and allocation decisions.
The proposed research will make essential contributions to the way CPS are implemented. This is a strategic area for Canada and Ontario, especially for its aerospace and automotive industry. A total of six HQPs will contribute to the research activities, and receive relevant training in the areas of computer architecture, operating systems and real-time analysis. My research will result in an open hardware / software platform based on the RISC-V architecture and fully demonstrated on FPGA. The on-going RISC-V revolution provides a unique opportunity to migrate developed techniques into industrial practice. The proposed solution will further lead to a new RISC-V extension.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
Ontario, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
RGPIN-2023-03749
Name nom
Pellizzoni, Rodolfo
Department département
Electrical and Computer Engineering
Organization id
33
Institution établissement
University of Waterloo
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2023
Award amount
52000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1507
Committee name en
Computer Science
Committee name fr
Sciences informatiques
Area of application code
802
Area of application group en
Information and communication services
Area of application group fr
Services d'information et de télécommunications
Area of application en
Information systems and technology
Area of application fr
Systèmes d'information et technologie de l'information
Research subject code
2702
Research subject group en
Information technology
Research subject group fr
Technologie de l'information
Research subject en
Computer systems organization
Research subject fr
Organisation des systèmes informatiques
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
A Hardware/Software Codesign Approach for Time-Predictable Computing
Keywords
embedded systems; real-time systems; operating systems; computer architecture; memory hierarchy
Application summary
Cyber-Physical Systems (CPS) are the next generation of smart, interconnected embedded systems. Many emerging CPS, such as unmanned aerial vehicles and autonomous cars, demand high computing performance for increased autonomy and real-time decision making, while meeting various constraints such as size, weight, power, and cost. CPS are safety-critical systems exhibiting real-time requirements: a failure to compute required actuation commands within a precise time window can lead to catastrophic results. For this reason, computing platforms used in CPS must be predictable, i.e., analyzable to derive tight bounds on worst-case task execution time. Unfortunately, existing multicore computing infrastructures are fundamentally inadequate to meet such timing constraints: their poor predictability - especially due to contention for access to shared hardware resources such as DRAM main memory and caches - is a major hurdle that makes verification and safety certification extremely challenging.
Existing research has tackled such predictability problem by introducing a set of software-based techniques to partition hardware resources between contending tasks or Virtual Machines (VMs). However, such approaches have fundamental limits as they do not directly influence resource allocation and arbitration decisions made at the hardware level. Consequently, predictability is typically achieved at the cost of performance degradation.
In this proposal, I argue that the key to maintaining predictability without performance compromises is a hardware / software codesign approach where hardware controllers expose suitable "control knobs" that can be tuned by the system software (OS / hypervisor) based on applications' requirements. At the hardware level, I will demonstrate the applicability of this concept by modifying both the shared cache and main memory controller, and building suitable control mechanisms in the CPU core. At the software level, I will design and implement suitable interfaces between applications, the OS and the hypervisor. Compared to a general-purpose system, in an embedded setting the system integrator typically has a precise knowledge of the applications' behavior, and such information can be exploited to optimize resource arbitration and allocation decisions.
The proposed research will make essential contributions to the way CPS are implemented. This is a strategic area for Canada and Ontario, especially for its aerospace and automotive industry. A total of six HQPs will contribute to the research activities, and receive relevant training in the areas of computer architecture, operating systems and real-time analysis. My research will result in an open hardware / software platform based on the RISC-V architecture and fully demonstrated on FPGA. The on-going RISC-V revolution provides a unique opportunity to migrate developed techniques into industrial practice. The proposed solution will further lead to a new RISC-V extension.