Chaotic dynamics in nanoscale NbO2 Mott memristors for analogue computing
Chaotic dynamics in nanoscale NbO2 Mott memristors for analogue computing
复制标题
DOI：
10.1038/nature23307
复制
发表时间：
2017-08-17
期刊：
影响因子：
64.8
通讯作者：
Williams, R. S. Tanley
中科院分区：
文献类型：
--
作者：
Kumar, Suhas;Strachan, John Paul;Williams, R. S. Tanley
目前，机器学习系统使用简化的神经元模型，缺乏生物系统中观察到的丰富的非线性现象，表现出时空合作动力学。有证据表明，神经元在一种被称为混沌边缘(1)的区域工作，这可能是大脑复杂性、学习效率、适应性和模拟(非布尔)计算的核心(2-7)。神经网络在混沌边缘运行时表现出更高的计算复杂性(2)，混沌元素网络已被提出用于解决组合或全局优化问题(8)。因此，可以结合到神经启发电路中的可控混沌行为的来源可能是未来计算系统的重要组成部分。这样的混沌元件已经使用复杂的晶体管电路来模拟，这种电路模拟了已知的混沌方程(9-12)，但还缺乏从单个可扩展电子设备实现混沌动力学的实验(5，6，13)。在这里，我们描述了每个直径小于100纳米的NbO2 Mott忆阻器，它们都表现出了非线性输运驱动的电流控制负阻和Mott转变驱动的温控负阻。Mott材料具有依赖于温度的金属-绝缘体转变，充当电子开关，从而在器件中引入依赖于历史的电阻。我们将这些忆阻器合并到弛豫振荡器(14)中，并观察到可调范围的周期和混沌自振荡(15)。结果表明，在纳米尺度上，非线性电流输运与热涨落耦合产生混沌振荡。这种忆阻器可以通过引入伪随机信号来防止全局同步，并在受限搜索期间帮助找到全局最小值，从而在某些类型的神经启发计算中有用。我们特别证明了将这种忆阻器结合到Hopfield计算网络的硬件中可以极大地提高收敛到计算困难问题的解的效率和精度。
At present, machine learning systems use simplified neuron models that lack the rich nonlinear phenomena observed in biological systems, which display spatio-temporal cooperative dynamics. There is evidence that neurons operate in a regime called the edge of chaos(1) that may be central to complexity, learning efficiency, adaptability and analogue (non-Boolean) computation in brains(2-7). Neural networks have exhibited enhanced computational complexity when operated at the edge of chaos(2), and networks of chaotic elements have been proposed for solving combinatorial or global optimization problems(8). Thus, a source of controllable chaotic behaviour that can be incorporated into a neural-inspired circuit may be an essential component of future computational systems. Such chaotic elements have been simulated using elaborate transistor circuits that simulate known equations of chaos(9-12), but an experimental realization of chaotic dynamics from a single scalable electronic device has been lacking(5,6,13). Here we describe niobium dioxide (NbO2) Mott memristors each less than 100 nanometres across that exhibit both a nonlinear-transport-driven current-controlled negative differential resistance and a Mott-transition-driven temperature-controlled negative differential resistance. Mott materials have a temperature-dependent metal-insulator transition that acts as an electronic switch, which introduces a history-dependent resistance into the device. We incorporate these memristors into a relaxation oscillator(14) and observe a tunable range of periodic and chaotic self-oscillations(15). We show that the nonlinear current transport coupled with thermal fluctuations at the nanoscale generates chaotic oscillations. Such memristors could be useful in certain types of neural-inspired computation by introducing a pseudo-random signal that prevents global synchronization and could also assist in finding a global minimum during a constrained search. We specifically demonstrate that incorporating such memristors into the hardware of a Hopfield computing network can greatly improve the efficiency and accuracy of converging to a solution for computationally difficult problems.