Memristive crossbars can efficiently implement Binarized Neural Networks (BNNs) wherein the weights are stored in high-resistance states (HRS) and low-resistance states (LRS) of the synapses. We propose SwitchX mapping of weights onto crossbars such that the power consumed by the crossbars and the impact of crossbar non-idealities, that lead to degradation in computational accuracy, are minimized. Essentially, SwitchX maps the binary weights in such manner that the crossbar comprises of more HRS than LRS synapses. Increased HRS in a crossbar will decrease the overall output dot-product current and thus lead to power savings. Interestingly, BNNs mapped onto crossbars with SwitchX also exhibit better robustness against adversarial attacks than the corresponding software BNN baseline as well as the standard crossbar mapped BNNs. Finally, we combine SwitchX with state-aware training (that further increases the feasibility of HRS states during weight mapping) to boost the robustness and energy-efficiency of BNN on hardware. We find that this approach yields stronger defense against adversarial attacks than Adversarial training, a state-of-the-art software defense. We perform experiments using benchmark datasets (CIFAR-100 & CIFAR-10) and show that SwitchX combined with state-aware training can yield upto ~35% improvements in clean accuracy and ~6-16% in adversarial accuracies against conventional BNNs on a 32x32 crossbar, while gaining ~22% savings in overall crossbar power consumption.

中文翻译：

---

SwitchX- Gmin-Gmax开关，用于忆阻Xbar上的高效节能和稳健的二进制神经网络实现

忆阻交叉开关可以有效地实现二进制神经网络（BNN），其中权重存储在突触的高电阻状态（HRS）和低电阻状态（LRS）中。我们提出将权重映射到交叉开关上的SwitchX，以使交叉开关所消耗的功率以及导致计算精度下降的交叉开关非理想性的影响最小。本质上，SwitchX以这样的方式映射二进制权重，即交叉开关包含的HRS比LRS突触更多。纵横制中HRS的增加将降低总的输出点积电流，从而节省功率。有趣的是，使用SwitchX映射到交叉开关上的BNN相对于相应的软件BNN基线以及标准的交叉开关映射的BNN，还表现出更好的抵抗攻击性的鲁棒性。最后，我们将SwitchX与状态感知训练相结合（这进一步增加了权重映射期间HRS状态的可行性），以提高BNN在硬件上的鲁棒性和能效。我们发现，与最先进的软件防御“对抗训练”相比，这种方法可产生更强大的防御对抗攻击。我们使用基准数据集（CIFAR-100和CIFAR-10）进行了实验，结果表明SwitchX与状态感知训练相结合，在32x32分辨率下，与传统的BNN相比，其清洁准确度提高了约35％，对抗精度提高了约6-16％。交叉开关，同时节省了约22％的总体交叉开关功耗。我们发现，与最先进的软件防御“对抗训练”相比，这种方法可产生更强大的防御对抗攻击。我们使用基准数据集（CIFAR-100和CIFAR-10）进行了实验，结果表明SwitchX与状态感知训练相结合，在32x32分辨率下，与传统的BNN相比，其清洁准确度提高了约35％，对抗精度提高了约6-16％。交叉开关，同时节省了约22％的总体交叉开关功耗。我们发现，与最先进的软件防御“对抗训练”相比，这种方法可产生更强大的防御对抗攻击。我们使用基准数据集（CIFAR-100和CIFAR-10）进行了实验，结果表明SwitchX与状态感知训练相结合，在32x32分辨率下，与传统的BNN相比，其清洁准确度提高了约35％，对抗精度提高了约6-16％。交叉开关，同时节省了约22％的总体交叉开关功耗。