Add to ORKGInternational audienceFor many types of integrated circuits, accepting larger failure rates in computations can be used to improve energy efficiency. We study the performance of faulty implementations of certain deep neural networks based on pessimistic and optimistic models of the effect of hardware faults. After identifying the impact of hyperparameters such as the number of layers on robustness, we study the ability of the network to compensate for computational failures through an increase of the network size. We show that some networks can achieve equivalent performance under faulty implementations, and quantify the required increase in computational complexity
As more deep learning algorithms enter safety-critical application domains, the importance of analyz...
Deep neural networks (DNNs) have been shown to tolerate “brain damage”: cumulative changes to the ne...
In the literature, it is argued that Deep Neural Networks (DNNs) possess a certain degree of robustn...
International audienceFor many types of integrated circuits, accepting larger failure rates in compu...
International audienceFor many types of integrated circuits, accepting larger failure rates in compu...
International audienceFor many types of integrated circuits, accepting larger failure rates in compu...
Deep Neural Networks (DNNs) are inherently computation-intensive and also power-hungry. Hardware acc...
With the development of neural networks based machine learning and their usage in mission critical a...
International audienceIn the literature, it is argued that Deep Neural Networks (DNNs) possess a cer...
International audienceIn the literature, it is argued that Deep Neural Networks (DNNs) possess a cer...
The entangled guardbands in terms of timing specification and energy budget ensure a system against ...
The entangled guardbands in terms of timing specification and energy budget ensure a system against ...
As computing systems continue their unquenchable rise towards and through million core architectures...
Wide attention was recently given to the problem of fault-tolerance in neural networks; while most a...
The recent success of deep neural networks (DNNs) in challenging perception tasks makes them a power...
As more deep learning algorithms enter safety-critical application domains, the importance of analyz...
Deep neural networks (DNNs) have been shown to tolerate “brain damage”: cumulative changes to the ne...
In the literature, it is argued that Deep Neural Networks (DNNs) possess a certain degree of robustn...
International audienceFor many types of integrated circuits, accepting larger failure rates in compu...
International audienceFor many types of integrated circuits, accepting larger failure rates in compu...
International audienceFor many types of integrated circuits, accepting larger failure rates in compu...
Deep Neural Networks (DNNs) are inherently computation-intensive and also power-hungry. Hardware acc...
With the development of neural networks based machine learning and their usage in mission critical a...
International audienceIn the literature, it is argued that Deep Neural Networks (DNNs) possess a cer...
International audienceIn the literature, it is argued that Deep Neural Networks (DNNs) possess a cer...
The entangled guardbands in terms of timing specification and energy budget ensure a system against ...
The entangled guardbands in terms of timing specification and energy budget ensure a system against ...
As computing systems continue their unquenchable rise towards and through million core architectures...
Wide attention was recently given to the problem of fault-tolerance in neural networks; while most a...
The recent success of deep neural networks (DNNs) in challenging perception tasks makes them a power...
As more deep learning algorithms enter safety-critical application domains, the importance of analyz...
Deep neural networks (DNNs) have been shown to tolerate “brain damage”: cumulative changes to the ne...
In the literature, it is argued that Deep Neural Networks (DNNs) possess a certain degree of robustn...