Neuroinspired unsupervised learning and pruning with subquantum CBRAM arrays
Neuroinspired unsupervised learning and pruning with subquantum CBRAM arrays
复制标题
DOI：
10.1038/s41467-018-07682-0
复制
发表时间：
2018-12-14
影响因子：
16.6
通讯作者：
Kuzum, Duygu
中科院分区：
文献类型：
--
作者：
Shi, Yuhan;Nguyen, Leon;Kuzum, Duygu
电阻式RAM交叉阵列提供了一种有吸引力的解决方案，以最大限度地减少片外数据传输和并行化神经网络的片上计算。在这里，我们报告了一个硬件/软件协同设计方法的基础上，低能量亚量子导电桥接RAM（CBRAM（R））设备和网络修剪技术，以减少网络级的能量消耗。首先，我们展示了低能量亚量子CBRAM器件，其具有渐进的开关特性，对于在无监督学习期间在硬件中实现权重更新非常重要。然后，我们开发了一个网络修剪算法，可以在训练过程中使用，不同于以前的网络修剪方法只适用于推理。使用一个512 kbit的子量子CBRAM阵列，我们实验证明了高识别精度的MNIST数据集的数字实现无监督学习。我们的硬件/软件协同设计方法可以为基于电阻记忆的神经启发系统铺平道路，这些系统可以在功率受限的情况下自主学习和处理信息。
Resistive RAM crossbar arrays offer an attractive solution to minimize off-chip data transfer and parallelize on-chip computations for neural networks. Here, we report a hardware/ software co-design approach based on low energy subquantum conductive bridging RAM (CBRAM (R)) devices and a network pruning technique to reduce network level energy consumption. First, we demonstrate low energy subquantum CBRAM devices exhibiting gradual switching characteristics important for implementing weight updates in hardware during unsupervised learning. Then we develop a network pruning algorithm that can be employed during training, different from previous network pruning approaches applied for inference only. Using a 512 kbit subquantum CBRAM array, we experimentally demonstrate high recognition accuracy on the MNIST dataset for digital implementation of unsupervised learning. Our hardware/software co-design approach can pave the way towards resistive memory based neuro-inspired systems that can autonomously learn and process information in power-limited settings.