Radiation-hardened
processors are planned to be supple against soft errors but such processors are
slower than Commercial Off-The-Shelf (COTS)processors as well significantly
costlier. In order to alleviate the high costs, software techniques such as task
re-executions must be organized together with adequately hardened
processors to provide reliability,this leads to a huge design space comprising
of the hardening level of the processors and the number of re-executions of
each task in the system. Each configuration in this design space represents a
tradeoff between processor load, reliability and costs.
The reliability comes at the price of higher costs due to higher levels of hardening and performance degradation due to hardening or due to re-executions.Thus, the tradeoffs between performance, reliability and costs must be carefully studied. Pertinent questions that arise in such a design scenario are — (i)how many times a task must be re-executed and (ii) what should be hardening level? — such that the system reliability is satisfied.
In order to evaluate such tradeoffs efficiently, in this project, we propose novel framework that harnesses the computational power of Graphics Processing Units (GPUs). Our framework is based on a system failure probability analysis that connects the probability of failure of tasks to the overall system reliability. Based on characteristics of this probabilistic analysis as well as real-time deadlines, we derive bounds on the design space to prune in feasible solutions. Finally, we illustrate the benefits of our proposed framework with several experiments
The reliability comes at the price of higher costs due to higher levels of hardening and performance degradation due to hardening or due to re-executions.Thus, the tradeoffs between performance, reliability and costs must be carefully studied. Pertinent questions that arise in such a design scenario are — (i)how many times a task must be re-executed and (ii) what should be hardening level? — such that the system reliability is satisfied.
In order to evaluate such tradeoffs efficiently, in this project, we propose novel framework that harnesses the computational power of Graphics Processing Units (GPUs). Our framework is based on a system failure probability analysis that connects the probability of failure of tasks to the overall system reliability. Based on characteristics of this probabilistic analysis as well as real-time deadlines, we derive bounds on the design space to prune in feasible solutions. Finally, we illustrate the benefits of our proposed framework with several experiments
Authors
Alhowaidi, Mohammad
Posts
0 comments:
Post a Comment