Glen Rosendale | Searcharxiv

arXiv2022

Universal Non-Polar Switching in Carbon-doped Transition Metal Oxides (TMOs) and Post TMOs

Transition metal oxides (TMOs) and post-TMOs (PTMOs), when doped with Carbon, show non-volatile current-voltage (I-V) characteristics, which are both universal and repeatable. We have shown spectroscopic evidence of the introduction of carbon-based impurity states inside the existing larger bandgap effectively creating a smaller bandgap which we suggest could enable Mott-like correlation effect. Our findings indicate new insights for yet to be understood unipolar and nonpolar resistive switching in the TMOs and PTMOs. We have shown that device switching is not thermal-energy dependent and have developed an electronic-dominated switching model that allows for the extreme temperature operation (from 1.5 K to 423 K) and state retention up to 673 K for a 1-hour bake. Importantly, we have optimized the technology in an industrial process and demonstrated integrated 1-transistor/1-resistor (1T1R) arrays up to 1 kbit with 47 nm devices on 300 mm wafers for advanced node CMOS-compatible correlated electron RAM (CeRAM). These devices are shown to operate with 2 ns write pulses and retain the memory states up to 200 C for 24 hours. The collection of attributes shown, including scalability to state-of-the-art dimensions, non-volatile operation to extreme low and high temperatures, fast write, and reduced stochasticity as compared to filamentary memories such as ReRAMs show the potential for a highly capable two-terminal back-end-of-line non-volatile memory.

cond-mat.mes-hall↗

arXiv2020

Training DNN IoT Applications for Deployment On Analog NVM Crossbars

A trend towards energy-efficiency, security and privacy has led to a recent focus on deploying DNNs on microcontrollers. However, limits on compute and memory resources restrict the size and the complexity of the ML models deployable in these systems. Computation-In-Memory architectures based on resistive nonvolatile memory (NVM) technologies hold great promise of satisfying the compute and memory demands of high-performance and low-power, inherent in modern DNNs. Nevertheless, these technologies are still immature and suffer from both the intrinsic analog-domain noise problems and the inability of representing negative weights in the NVM structures, incurring in larger crossbar sizes with concomitant impact on ADCs and DACs. In this paper, we provide a training framework for addressing these challenges and quantitatively evaluate the circuit-level efficiency gains thus accrued. We make two contributions: Firstly, we propose a training algorithm that eliminates the need for tuning individual layers of a DNN ensuring uniformity across layer weights and activations. This ensures analog-blocks that can be reused and peripheral hardware substantially reduced. Secondly, using NAS methods, we propose the use of unipolar-weighted (either all-positive or all-negative weights) matrices/sub-matrices. Weight unipolarity obviates the need for doubling crossbar area leading to simplified analog periphery. We validate our methodology with CIFAR10 and HAR applications by mapping to crossbars using 4-bit and 2-bit devices. We achieve up to 92:91% accuracy (95% floating-point) using 2-bit only-positive weights for HAR. A combination of the proposed techniques leads to 80% area improvement and up to 45% energy reduction.

cs.LG↗