Evaluation and Mitigation of Soft-Errors in Neural Network-Based Object Detection in Three GPU Architectures

Santos, Fernando Fernandes dos; Draghetti, Lucas Klein; Weigel, Lucas; Carro, Luigi; Navaux, Philippe O. A.; Rech, Paolo

doi:10.1109/dsn-w.2017.47

Cited by 42 publications

(29 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the literature, several works have been proposed in the last years for addressing the reliability assessment of CNNs. Concerning physical-based FI, we can cite [2], where the reliability dependence on three different Graphics Processing Unit (GPU) architectures (Kepler, Maxwell, and Pascal) was evaluated executing the Darknet Neural Network [8] when exposed to atmospheric-like neutrons. In [9], the authors analyse the reliability of a 54-layer Deep Neural Network injecting faults in the network weights and input data using an accelerated neutron beam for studying transient errors and FI tests to simulate permanent faults.…”

Section: Related Workmentioning

confidence: 99%

“…Nowadays, there is an incentive for pushing CNNs from the cloud to the edge devices and, in particular, for real-time safety-critical systems, e.g., in autonomous driving. Several recent studies have demonstrated that hardware faults induced by external perturbations (i.e., in a harsh environment) can significantly impact the inference leading to CNN prediction failures [2], [3]. Therefore, ensuring the reliability of CNNs is crucial, especially when deployed in safety-and mission-critical applications, such as robotics, aeronautics, smart healthcare, and autonomous driving.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Emulating the Effects of Radiation-Induced Soft-Errors for the Reliability Assessment of Neural Networks

Matanaluza

Ruospo

Söderström³

et al. 2022

IEEE Trans. Emerg. Topics Comput.

View full text Add to dashboard Cite

Convolutional Neural Networks (CNNs) are currently one of the most widely used predictive models in machine learning. Recent studies have demonstrated that hardware faults induced by radiation fields, including cosmic rays, may significantly impact the CNN inference leading to wrong predictions. Therefore, ensuring the reliability of CNNs is crucial, especially for safety-critical systems. In the literature, several works propose reliability assessments of CNNs mainly based on statistically injected faults. This work presents a software emulator capable of injecting real faults retrieved from radiation tests. Specifically, from the device characterisation of a DRAM memory, we extracted event rates and fault models. The software emulator can reproduce their incidence and access their effect on CNN applications with a reliability assessment precision close to the physical one. Radiation-based physical injections and emulator-based injections are performed on three CNNs (LeNet-5) exploiting different data representations. Their outcomes are compared, and the software results evidence that the emulator is able to reproduce the faulty behaviours observed during the radiation tests for the targeted CNNs. This approach leads to a more concise use of radiation experiments since the extracted fault models can be reused to explore different scenarios (e.g., impact on a different application).

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Emulating the Effects of Radiation-Induced Soft-Errors for the Reliability Assessment of Neural Networks

Matanaluza

Ruospo

Söderström³

et al. 2022

IEEE Trans. Emerg. Topics Comput.

View full text Add to dashboard Cite

show abstract

“…Although several studies [17][8] [18][19] [18] have evaluated and analyzed the reliability of DNN models, and many techniques have been proposed to mitigation the soft errors in GPUs based on software solutions. For instance, Triple Modular Redundancy (TMR), Double Modular Redundancy (DMR), and Algorithm-Based Fault Tolerance (ABFT).…”

Section: Introductionmentioning

confidence: 99%

A Selective Mitigation Technique of Soft Errors for DNN Models Used in Healthcare Applications: DenseNet201 Case Study

2021

View full text Add to dashboard Cite

Deep neural networks (DNNs) have been successfully deployed in widespread domains, including healthcare applications. DenseNet201 is a new DNN architecture used in healthcare systems (i.e., presence detection of the surgical tool). Specialized accelerators such as GPUs have been used to speed up the execution of DNNs. Nevertheless, GPUs are prone to transient effects and other reliability threats, which can impact DNN models' reliability. Safety-critical systems, such as healthcare applications, must be highly reliable because minor errors might lead to severe injury or death. In this paper, we propose a selective mitigation technique that relies on in-depth analysis. First, we inject the DenseNet201 model implemented on a GPU via NVIDIA's SASSIFI fault injector. Second, we perform a comprehensive analysis from the perspective of kernel and layer to identify the most vulnerable portions of the injected model. Finally, we validate our technique by applying it to the top-vulnerable kernels to selectively protect the only sensitive portions of the model to avoid unnecessary overheads. Our experiments demonstrate that our mitigation technique achieves a significant reduction in the percentage of errors that cause malfunction (errors that lead to misclassification) from 6.463% to 0.21%. Moreover, the performance overhead (the execution time) of our technique is compared with the well-known protection techniques: Algorithm-Based Fault Tolerance (ABFT), Double Modular Redundancy (DMR), and Triple Modular Redundancy (TMR). The proposed solution shows only 0.3035% overhead compared to these techniques while correcting up 84.8% of the SDC errors in DenseNet201, remarkably improving the healthcare domain's model reliability.

show abstract

“…However, TMR introduces large overhead in hardware design, especially for massively parallel processing array. Moreover, ECC can only be used in a memory unit and not in a logic unit owing to its inability to maintain the error correction message transfer between the layers [4]. For its part, fault-aware training has been tested and found to be effective for many DNN inference accelerator fault-patterns, while also having little impact on hardware design.…”

Section: Introductionmentioning

confidence: 99%

ACIMS: Analog CIM Simulator for DNN Resilience

Dong¹,

Wang²,

Yang³

et al. 2021

Electronics

View full text Add to dashboard Cite

Analog Computing In Memory (ACIM) combines the advantages of both Compute In Memory (CIM) and analog computing, making it suitable for the design of energy-efficient hardware accelerators for computationally intensive DNN applications. However, their use will introduce hardware faults that decrease the accuracy of DNN. In this work, we take Sandwich-Ram as the real hardware example of ACIM and are the first to propose a fault injection and fault-aware training framework for it, named Analog Computing In Memory Simulator (ACIMS). Using this framework, we can simulate and repair the hardware faults of ACIM. The experimental results show that ACIMS can recover 91.0%, 93.7% and 89.8% of the DNN’s accuracy drop through retraining on the MNIST, SVHN and Cifar-10 datasets, respectively; moreover, their adjusted accuracy can reach 97.0%, 95.3% and 92.4%.

show abstract

Evaluation and Mitigation of Soft-Errors in Neural Network-Based Object Detection in Three GPU Architectures

Cited by 42 publications

References 33 publications

Emulating the Effects of Radiation-Induced Soft-Errors for the Reliability Assessment of Neural Networks

Emulating the Effects of Radiation-Induced Soft-Errors for the Reliability Assessment of Neural Networks

A Selective Mitigation Technique of Soft Errors for DNN Models Used in Healthcare Applications: DenseNet201 Case Study

ACIMS: Analog CIM Simulator for DNN Resilience

Contact Info

Product

Resources

About