Benchmarking Inference Performance of Deep Learning Models on Analog Devices
This research provides insights into the robustness of deep learning models to noise on analog hardware, which is crucial for developers designing efficient edge computing devices.
This study systematically evaluated the inference performance of popular image classification models deployed on analog devices by adding additive white Gaussian noise to the weights during inference. It found that deeper models and those with more redundancy, like VGG, are generally more robust to noise, though performance is also influenced by model design, structure, task, and dataset.
Analog hardware implemented deep learning models are promising for computation and energy constrained systems such as edge computing devices. However, the analog nature of the device and the associated many noise sources will cause changes to the value of the weights in the trained deep learning models deployed on such devices. In this study, systematic evaluation of the inference performance of trained popular deep learning models for image classification deployed on analog devices has been carried out, where additive white Gaussian noise has been added to the weights of the trained models during inference. It is observed that deeper models and models with more redundancy in design such as VGG are more robust to the noise in general. However, the performance is also affected by the design philosophy of the model, the detailed structure of the model, the exact machine learning task, as well as the datasets.