Xuefeng Gu

Xuefeng Gu

Auto-generated profile, based on publicly available information

Xuefeng Gu

Center for Heterogeneous Integration and Performance Scaling (CHIPS), Department of Electrical and Computer Engineering, University of California, Los Angeles, Los Angeles, CA 90095 USA

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
24 July 2018 in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
Verified
Journal Article
Closed access
YDYuan Du
Yuan Du
LDLi Du
Li Du
XGXuefeng Gu
Xuefeng Gu
JDJieqiong Du
Jieqiong Du
XWX. Shawn Wang
X. Shawn Wang
MCMau-Chung Frank Chang
Mau-Chung Frank Chang
An analog neural network computing engine based on CMOS-compatible charge-trap transistor (CTT) is proposed in this paper. CTT devices are used as analog multipliers. Compared to digital multipliers, CTT-based analog multiplier shows significant area and power reduction. The proposed computing engine is composed of a scalable CTT multiplier array and energy efficient analog-digital interfaces. By implementing the sequential analog fabric (SAF), the engine’s mixed-signal interfaces are simplified and hardware overhead remains constant regardless of the size of the array. A proof-of-concept 784 by 784 CTT computing engine is implemented using TSMC 28nm CMOS technology and occupies 0.68mm². The simulated performance achieves 76.8 TOPS (8-bit) with 500 MHz clock frequency and consumes 14.8 mW. As an example, we utilize this computing engine to address a classic pattern recognition problem – classifying handwritten digits on MNIST database and obtained a performance comparable to state-of-the-art fully connected neural networks using 8-bit fixed-point resolution.

4 July 2017 in IEEE Electron Device Letters
Verified
Journal Article
Closed access
XGXuefeng Gu
Xuefeng Gu
SISubramanian S. Iyer
Subramanian S. Iyer
Unsupervised learning is demonstrated using a device ubiquitously found in today's technology: a transistor with high-k -metal gate. Specifically, the charge-trapping phenomenon in the high-k gate dielectric is leveraged so that the device can be used as a non-volatile analog memory. Experimental data from 22-nm silicon-on-insulator devices reveal that a charge-trap transistor possesses promising characteristics for implementing synapses in neural networks, such as very fine tunability, weight-dependent plasticity, and low power consumption. A proof-of-concept winner-takes-all neural network is simulated based on experimental data and perfect clustering is achieved within tens of training cycles. This means that the network can be trained for multiple times, and a larger system can be built. The robustness of the procedure to the device variation is also discussed.

15 August 2019 in IEEE Transactions on Electron Devices
Verified
Journal Article
Closed access
XGXuefeng Gu
Xuefeng Gu
ZWZhe Wan
Zhe Wan
SISubramanian S. Iyer
Subramanian S. Iyer
Since our demonstration of unsupervised learning using the CMOS-only charge-trap transistors (CTTs) as analog synapses, there has been an increasing interest in exploiting the device for various other neural network (NN) applications. However, most of these studies are limited to mere simulation due to the absence of detailed experimental device characterization. In this article, we provide a comprehensive investigation of the programming behavior of CTTs, including analog retention, intra- and inter-device variation, and fine-tuning of the device, both for individual devices and for devices in an integrated array. It is found that, after programming, the channel current gradually increases to a higher level, and the shift is larger when the device is programmed to a higher threshold voltage. With this postprogramming current increase appropriately accounted for, individual devices can be programmed to an equivalent precision of five bits, and three bits can be achieved for devices in an array. Our results reveal the promising future of using the CTT as a CMOS-only analog memory device.