ECRAM as Scalable Synaptic Cell for High-Speed, Low-Power Neuromorphic Computing
| Autor: | Tayfun Gokmen, Douglas M. Bishop, Seyoung Kim, Matthew Copel, Paul M. Solomon, Sanghoon Shin, Jianshi Tang, Teodor K. Todorov, Kevin K. Chan, K.-L. Lee, John Rozen, Wilfried Haensch |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
010302 applied physics
Physics business.industry chemistry.chemical_element Conductance Charge (physics) 02 engineering and technology 021001 nanoscience & nanotechnology 01 natural sciences Power (physics) Amplitude chemistry Neuromorphic engineering 0103 physical sciences Optoelectronics Lithium 0210 nano-technology business Energy (signal processing) Pulse-width modulation |
| Zdroj: | 2018 IEEE International Electron Devices Meeting (IEDM). |
| Popis: | We demonstrate a nonvolatile Electro-Chemical Random-Access Memory (ECRAM) based on lithium (Li) ion intercalation in tungsten oxide (WO 3 ) for high-speed, low-power neuromorphic computing. Symmetric and linear update on the channel conductance is achieved using gate current pulses, where up to 1000 discrete states with large dynamic range and good retention are demonstrated. MNIST simulation based on the experimental data shows an accuracy of 96%. For the first time, high-speed programming with pulse width down to 5 ns and device operation at scales down to $300\times 300\ \text{nm}^{2}$ are shown, confirming the technological relevance of ECRAM for neuromorphic array implementation. It is also verified that the conductance change scales linearly with pulse width, amplitude and charge, projecting an ultralow switching energy ∼1 fJ for $100\times 100\ \text{nm}^{2}$ devices. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|