Wenhao Song

Wenhao Song

Auto-generated profile, based on publicly available information

Wenhao Song

University of Southern California, Los Angeles, CA 90089, USA

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

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

Representative Publications

View All Publications
29 March 2023 in Nature
Verified
Journal Article
Closed access
PDF file
MRMingyi Rao
Mingyi Rao
HTHao Tang
Hao Tang
JWJiangbin Wu
Jiangbin Wu
WSWenhao Song
Wenhao Song
MZMax Zhang
Max Zhang
JYJ. Joshua Yang
J. Joshua Yang

26 June 2020 in Science Advances
Verified
Journal Article
Open access
PDF file
CWChen-Yu Wang
Chen-Yu Wang
SLShi-Jun Liang
Shi-Jun Liang
SWShuang Wang
Shuang Wang
PWPengfei Wang
Pengfei Wang
ZLZhu’an Li
Zhu’an Li
FMFeng Miao
Feng Miao
A reconfigurable neural network vision sensor is proposed by using gate-tunable photoresponse of van der Waals heterostructures.

23 February 2024 in Science
Verified
Journal Article
Closed access
WSWenhao Song
Wenhao Song
MRMingyi Rao
Mingyi Rao
YLYunning Li
Yunning Li
CLCan Li
Can Li
YZYe Zhuo
Ye Zhuo
JYJ. Joshua Yang
J. Joshua Yang
In-memory computing represents an effective method for modeling complex physical systems that are typically challenging for conventional computing architectures but has been hindered by issues such as reading noise and writing variability that restrict scalability, accuracy, and precision in high-performance computations. We propose and demonstrate a circuit architecture and programming protocol that converts the analog computing result to digital at the last step and enables low-precision analog devices to perform high-precision computing. We use a weighted sum of multiple devices to represent one number, in which subsequently programmed devices are used to compensate for preceding programming errors. With a memristor system-on-chip, we experimentally demonstrate high-precision solutions for multiple scientific computing tasks while maintaining a substantial power efficiency advantage over conventional digital approaches.

25 July 2025 in Science Advances
Verified
Journal Article
Open access
ZWZixu Wang
Zixu Wang
WSWenhao Song
Wenhao Song
TWTong Wang
Tong Wang
ZWZihan Wang
Zihan Wang
YXYichun Xu
Yichun Xu
JYJ. Joshua Yang
J. Joshua Yang
The von Neumann bottleneck has led to a substantial rise in energy consumption of computing hardware and memory systems, particularly for data-intensive tasks like signal processing. Memristor-based in-memory computing offers an efficient alternative by performing computations within analog memory. Here, we demonstrate real-time signal processing using a fused network that combines the real-time discrete Fourier transform (DFT) and convolutional neural network (CNN) on a memristor-based analog system on a chip (SoC). A 128-by-128 memristor crossbar array performs the DFT on audio signals with a peak signal-to-noise ratio of 33.49 dB, while the following CNN classifies the resulting spectrograms with 94.72% accuracy on the AudioMNIST dataset. In addition, convolution-based edge detection is applied to real-time video frames. The SoC offers substantial energy efficiency improvement over traditional digital systems in signal processing tasks. This work highlights the potential of memristor-based SoCs for efficient real-time signal processing.

14 March 2022
Verified
Conference Paper
Closed access
PDF file
LZLeilei Zhu
Leilei Zhu
XSXin Sui
Xin Sui
WSWenhao Song
Wenhao Song
RLRuixiang Liu
Ruixiang Liu
DLDan Liu
Dan Liu
LLLi Li
Li Li
With the continuous expansion of cloud data center scale, the flow of network traffic generated by virtual machine scheduling also increased, resulting in a sharp increase of energy consumption in cloud data center. Under certain network topology, different virtual machine scheduling strategies would produce different network flow. Therefore, how to optimize the scheduling strategy and reduce network flow became the key factors to reduce the energy consumption of cloud datacenter. In this paper, based on the cloud computing simulation platform CloudSim, the network module was extended, adding the network communication module of virtual machine migrating. In order to ensure the stability of the network data center, a virtual machine scheduling algorithm NPABFD for network traffic and energy consumption aware was proposed. The experimental results showed that the NPABFD algorithm could effectively reduce the transfer traffic of migrating VMs in high-level switches, reduce the global NDC's global energy consumption and the migrating number of VMs.