High-end embedded systems such as smart phones, game consoles, GPS-enabled automotive systems, and home entertainment centers, are becoming ubiquitous. Like their general-purpose counterparts, and for many of the same energy-related reasons, embedded sys- tems are turning to multicore architectures. Moreover, as the de- mand for more compute-intensive capabilities for embedded sys- tems increases, these multicore architectures will evolve into many- core systems for improved performance or performance/area/Watt. These systems are often organized as cluster based Non-Uniform Memory Access (NUMA) architectures that provide the program- mer with a shared-memory abstraction, with the cost of sharing memory (in terms of performance, energy, and complexity) varying substantially depending on the locations of the communicating pro- cesses. This paper investigates one of the principal challenges pre- sented by these emerging NUMA architectures for embedded sys- tems: providing efficient, energy-effective and convenient mecha- nisms for synchronization and communication. In this paper, we propose an initial solution based on hardware support for specula- tive synchronization.
Dimitra Papagiannopoulou, R. Iris Bahar, Tali Moreshet, Maurice Herlihy, Andrea Marongiu, Luca Benini (2013). Transparent and energy-efficient speculation on NUMA architectures for embedded MPSoCsProceedings of the First International Workshop on Many-core Embedded Systems - MES '13. New York : ACM [10.1145/2489068.2489078].
Transparent and energy-efficient speculation on NUMA architectures for embedded MPSoCsProceedings of the First International Workshop on Many-core Embedded Systems - MES '13
MARONGIU, ANDREA;BENINI, LUCA
2013
Abstract
High-end embedded systems such as smart phones, game consoles, GPS-enabled automotive systems, and home entertainment centers, are becoming ubiquitous. Like their general-purpose counterparts, and for many of the same energy-related reasons, embedded sys- tems are turning to multicore architectures. Moreover, as the de- mand for more compute-intensive capabilities for embedded sys- tems increases, these multicore architectures will evolve into many- core systems for improved performance or performance/area/Watt. These systems are often organized as cluster based Non-Uniform Memory Access (NUMA) architectures that provide the program- mer with a shared-memory abstraction, with the cost of sharing memory (in terms of performance, energy, and complexity) varying substantially depending on the locations of the communicating pro- cesses. This paper investigates one of the principal challenges pre- sented by these emerging NUMA architectures for embedded sys- tems: providing efficient, energy-effective and convenient mecha- nisms for synchronization and communication. In this paper, we propose an initial solution based on hardware support for specula- tive synchronization.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.