论文标题
铁氟烷中内存计算的证据
Evidence of In-Memory Computing in a Ferrofluid
论文作者
论文摘要
磁性流体是重要研究领域的出色候选者，包括能量收集，生物医学应用，软机器人和勘探。但是，尽管有相关的进步，例如形状可重构性，但没有证据表明其计算能力，包括仿真突触功能。在这里，我们在实验上证明了FE3O4水基铁氟烷（FF）可以执行电气模拟计算并使用准直流信号进行编程，并以射频（RF）模式读取。我们已经在各个方面都观察到归因于回忆行为的特征，这些特征具有短期和长期信息存储能力和可塑性。胶体能够使用自定义内存信号处理方案以及通过训练读出层来对8x8像素数据集的数字进行分类。
Magnetic fluids are excellent candidates for important research fields including energy harvesting, biomedical applications, soft robotics and exploration. However, notwithstanding relevant advancements such as shape reconfigurability, that have been demonstrated, there is no evidence for their computation capability, including the emulation of synaptic functions. Here, we experimentally demonstrate that a Fe3O4 water-based Ferrofluid (FF) can perform electrical analog computing and be programmed using quasi DC signals and read at Radio Frequency (RF) mode. We have observed features in all respects attributable to a memristive behavior, featuring both short and long-term information storage capacity and plasticity. The colloid was capable of classifying digits of a 8x8 pixel dataset using a custom in-memory signal processing scheme, and through Physical Reservoir Computing (PRC) by training a readout layer.
