Input-Dependent Edge-Cloud Mapping of Recurrent Neural Networks Inference
Input-Dependent Edge-Cloud Mapping of Recurrent Neural Networks Inference
复制标题
DOI：
10.1109/dac18072.2020.9218595
复制
发表时间：
2020-07
期刊：
影响因子：
--
通讯作者：
D. J. Pagliari;R. Chiaro;Yukai Chen;S. Vinco;E. Macii;M. Poncino
中科院分区：
文献类型：
--
作者：
D. J. Pagliari;R. Chiaro;Yukai Chen;S. Vinco;E. Macii;M. Poncino
鉴于递归神经网络(RNN)推理的计算复杂性，物联网和移动设备通常会将此任务卸载到云中。然而，RNN推理的执行时间和能量消耗强烈地依赖于处理的输入的长度。因此，考虑到通信成本，在本地处理短输入序列可能会更方便，而只将长输入序列卸载到云中。在本文中，我们提出了一个低开销的运行时工具，它可以自动执行这个选择。基于真实边缘和云设备的结果表明，与完全在本地或完全在云中运行RNN推理的解决方案相比，我们的方法能够同时减少系统的总执行时间和能量消耗。
Given the computational complexity of Recurrent Neural Networks (RNNs) inference, IoT and mobile devices typically offload this task to the cloud. However, the execution time and energy consumption of RNN inference strongly depends on the length of the processed input. Therefore, considering also communication costs, it may be more convenient to process short input sequences locally and only offload long ones to the cloud. In this paper, we propose a low-overhead runtime tool that performs this choice automatically. Results based on real edge and cloud devices show that our method is able to simultaneously reduce the total execution time and energy consumption of the system compared to solutions that run RNN inference fully locally or fully in the cloud.