Reservoir Computing for Temporal Data Classification Using a Dynamic Solid Electrolyte ZnO Thin Film Transistor

Abstract

The processing of sequential and temporal data is essential to computer vision and speech recognition, two of the most common applications of artificial intelligence (AI). Reservoir computing (RC) is a branch of AI that offers a highly efficient framework for processing temporal inputs at a low training cost compared to conventional Recurrent Neural Networks (RNNs). However, despite extensive effort, two-terminal memristor-based reservoirs have, until now, been implemented to process sequential data by reading their conductance states only once, at the end of the entire sequence. This method reduces the dimensionality, related to the number of signals from the reservoir and thereby lowers the overall performance of reservoir systems. Higher dimensionality facilitates the separation of originally inseparable inputs by reading out from a larger set of spatiotemporal features of inputs. Moreover, memristor-based reservoirs either use multiple pulse rates, fast or slow read (immediately or with a delay introduced after the end of the sequence), or excitatory pulses to enhance the dimensionality of reservoir states. This adds to the complexity of the reservoir system and reduces power efficiency. In this paper, we demonstrate the first reservoir computing system based on a dynamic three terminal solid electrolyte ZnO/Ta2O5 Thin-film Transistor fabricated at less than 100°C. The inherent nonlinearity and dynamic memory of the device lead to a rich separation property of reservoir states that results in, to our knowledge, the highest accuracy of 94.44%, using electronic charge-based system, for the classification of hand-written digits. This improvement is attributed to an increase in the dimensionality of the reservoir by reading the reservoir states after each pulse rather than at the end of the sequence. The third terminal enables a read operation in the off state, that is when no pulse is applied at the gate terminal, via a small read pulse at the drain. This fundamentally allows multiple read operations without increasing energy consumption, which is not possible in the conventional two-terminal memristor counterpart. Further, we have also shown that devices do not saturate even after multiple write pulses which demonstrates the device’s ability to process longer sequences.

Leaky FinFET for Reservoir Computing with Temporal Signal Processing

“…However, training of the readout network connected to the reservoirs can allow them to be differentiated for a reservoir computing system. 14 Device-to-device and cycle-to-cycle variations of the L-FinFET reservoir operations were confirmed, as shown in Figure S4 of the Supporting Information. Although there are some variations, most of them are not significant enough to overturn the reservoir states.…”

“…Although most of the reservoir states were distinguished, some of them appeared visually in close proximity to each other. However, training of the readout network connected to the reservoirs can allow them to be differentiated for a reservoir computing system . Device-to-device and cycle-to-cycle variations of the L-FinFET reservoir operations were confirmed, as shown in Figure S4 of the Supporting Information.…”

“…Recently, a three-terminal-based physical reservoir was developed to reduce the power consumption further through the use of a gate terminal to enable multiple write operations in the off-state. As an example, a thin-film transistor with a solid-state electrolyte was used . However, its large-sized footprint area with a channel width of 100 μm and a channel length of 1.5 μm is a concern in terms of scalability compared to a memristor-based reservoir.…”

“…As an example, a thin-film transistor with a solid-state electrolyte was used. 14 However, its large-sized footprint area with a channel width of 100 μm and a channel length of 1.5 μm is a concern in terms of scalability compared to a memristor-based reservoir.…”

Semiconductor technologies and related topics for implementation of electronic reservoir computing systems

“…Si-based PRC using a disordered dopant-atom network [27] is a unique system exploiting the variation in material level for computation. As PRC systems harnessing other non-Si CMOS materials, ZnO [28] and organic semiconductor [29] have been reported recently. Both use electrolytes for gating to exploit the difference in time scale of dynamics between ions and electrons/holes.…”