TaxoNN: A Light-Weight Accelerator for Deep Neural Network Training
This work addresses the need for energy-efficient DNN training on embedded devices, offering incremental improvements in hardware design.
The paper tackles the problem of enabling efficient deep neural network training on embedded devices by proposing TaxoNN, a light-weight accelerator that reuses inference hardware with low-bitwidth units, achieving a 0.97% higher misclassification rate compared to full-precision, along with 2.1x power saving and 1.65x area reduction over state-of-the-art.
Emerging intelligent embedded devices rely on Deep Neural Networks (DNNs) to be able to interact with the real-world environment. This interaction comes with the ability to retrain DNNs, since environmental conditions change continuously in time. Stochastic Gradient Descent (SGD) is a widely used algorithm to train DNNs by optimizing the parameters over the training data iteratively. In this work, first we present a novel approach to add the training ability to a baseline DNN accelerator (inference only) by splitting the SGD algorithm into simple computational elements. Then, based on this heuristic approach we propose TaxoNN, a light-weight accelerator for DNN training. TaxoNN can easily tune the DNN weights by reusing the hardware resources used in the inference process using a time-multiplexing approach and low-bitwidth units. Our experimental results show that TaxoNN delivers, on average, 0.97% higher misclassification rate compared to a full-precision implementation. Moreover, TaxoNN provides 2.1$\times$ power saving and 1.65$\times$ area reduction over the state-of-the-art DNN training accelerator.