Fuxi Cai
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Fuxi Cai has a total of 92 co-authors. Here are the top co-authors by number of shared publications:
- Wei Lu
Department of Electrical Engineering and Computer Science, University of Michigan, Ann Arbor, USA
7 shared publications Wei D. Lu
University of Michigan, Ann Arbor, MI 48109 USA
6 shared publicationsWen Ma
Department of Electrical Engineering and Computer Science, University of Michigan, Ann Arbor, MI, USA
5 shared publicationsChao Du
Department of Electrical Engineering and Computer Science, University of Michigan, Ann Arbor, MI, USA
4 shared publications- Seung Hwan Lee
Department of Electronic Engineering Kyung Hee University Yongin South Korea
4 shared publications - Jianhua Joshua Yang
Ming Hsieh Department of Electrical and Computer Engineering University of Southern California Los Angeles California USA
3 shared publications
Latest Publications
- Open Access
- 25 July 2025
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 11 (30) , eadv3436
Real-time signal processing enabled by fused networks on a memristor-based system on a chip
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.
- 25 July 2025
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 11 (30) , eadv3436
Open AccessReal-time signal processing enabled by fused networks on a memristor-based system on a chip
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.
- Top Cited Papers
- 23 February 2024
- Published by American Association for the Advancement of Science (AAAS) in Science
- Vol. 383 (6685) , 903-910
Programming memristor arrays with arbitrarily high precision for analog computing
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.
- 23 February 2024
- Published by American Association for the Advancement of Science (AAAS) in Science
- Vol. 383 (6685) , 903-910
Top Cited PapersProgramming memristor arrays with arbitrarily high precision for analog computing
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.
- Open Access
- 1 May 2022
- Published by Wiley in Advanced Intelligent Systems
- Vol. 4 (8)
A Fully Integrated System‐on‐Chip Design with Scalable Resistive Random‐Access Memory Tile Design for Analog in‐Memory Computing
As the demands of big data applications and deep learning continue to rise, the industry is increasingly looking to artificial intelligence (AI) accelerators. Analog in‐memory computing (AiMC) with emerging nonvolatile devices enable good hardware solutions, due to its high energy efficiency in accelerating the multiply‐and‐accumulation (MAC) operation. Herein, an Applied Materials custom‐designed system‐on‐chip (SoC) targeting AI applications with analog in‐memory computing using resistive random‐access memory (ReRAM) as the compute element is demonstrated. The first silicon achieves high energy efficiency in MAC operations. This chip is implemented with LeNet‐1 neural network on ReRAM tiles and demonstrated by Modified National Institute of Standards and Technology (MNIST) classification with accuracy matching that predicted in the simulations. A simulation framework, AI Sim, is also developed to evaluate the system performance for large‐scale application and guide the bitcell development and design choices.
- 1 May 2022
- Published by Wiley in Advanced Intelligent Systems
- Vol. 4 (8)
Open AccessA Fully Integrated System‐on‐Chip Design with Scalable Resistive Random‐Access Memory Tile Design for Analog in‐Memory Computing
As the demands of big data applications and deep learning continue to rise, the industry is increasingly looking to artificial intelligence (AI) accelerators. Analog in‐memory computing (AiMC) with emerging nonvolatile devices enable good hardware solutions, due to its high energy efficiency in accelerating the multiply‐and‐accumulation (MAC) operation. Herein, an Applied Materials custom‐designed system‐on‐chip (SoC) targeting AI applications with analog in‐memory computing using resistive random‐access memory (ReRAM) as the compute element is demonstrated. The first silicon achieves high energy efficiency in MAC operations. This chip is implemented with LeNet‐1 neural network on ReRAM tiles and demonstrated by Modified National Institute of Standards and Technology (MNIST) classification with accuracy matching that predicted in the simulations. A simulation framework, AI Sim, is also developed to evaluate the system performance for large‐scale application and guide the bitcell development and design choices.
- Top Cited Papers
- 6 July 2020
- Published by Springer Nature in Nature Electronics
- Vol. 3 (7) , 409-418
Power-efficient combinatorial optimization using intrinsic noise in memristor Hopfield neural networks
- 6 July 2020
- Published by Springer Nature in Nature Electronics
- Vol. 3 (7) , 409-418
Top Cited PapersPower-efficient combinatorial optimization using intrinsic noise in memristor Hopfield neural networks
- Open Access
- 4 May 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Exploratory Solid-State Computational Devices and Circuits
- Vol. 6 (1) , 36-44
A Fully Integrated Reprogrammable CMOS-RRAM Compute-in-Memory Coprocessor for Neuromorphic Applications
Analog compute-in-memory with resistive random access memory (RRAM) devices promises to overcome the data movement bottleneck in data-intensive AI and machine learning. RRAM crossbar arrays improve the efficiency of vector-matrix multiplications (VMM), which is a vital operation in these applications. The prototype IC is the first complete, fully-integrated analog-RRAM CMOS coprocessor. This article focuses on the digital and analog circuitry that supports efficient and flexible RRAM-based computation. A passive 54x108 RRAM crossbar array performs VMM in the analog domain. Specialized mixed-signal circuits stimulate and read the outputs of the RRAM crossbar. The single-chip CMOS prototype includes a RISC processor interfaced to a memory-mapped mixed-signal core. In the mixed-signal core, ADCs and DACs interface to the passive RRAM crossbar. The RISC processor controls the mixed-signal circuits and the algorithm data path. The system is fully programmable and supports forward and backward propagation. As proof of concept, a fully-integrated 0.18μm CMOS prototype with a post-processed RRAM array demonstrates several key functions of machine learning including online learning. The mixed-signal core consumes 64mW at an operating frequency of 148MHz. The total system power consumption, considering the mixed-signal circuitry, the digital processor, and the passive RRAM array is 307mW. The maximum theoretical throughput is 2.6 GOPS at an efficiency of 8.5 GOPS/W.
- 4 May 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Exploratory Solid-State Computational Devices and Circuits
- Vol. 6 (1) , 36-44
Open AccessA Fully Integrated Reprogrammable CMOS-RRAM Compute-in-Memory Coprocessor for Neuromorphic Applications
Analog compute-in-memory with resistive random access memory (RRAM) devices promises to overcome the data movement bottleneck in data-intensive AI and machine learning. RRAM crossbar arrays improve the efficiency of vector-matrix multiplications (VMM), which is a vital operation in these applications. The prototype IC is the first complete, fully-integrated analog-RRAM CMOS coprocessor. This article focuses on the digital and analog circuitry that supports efficient and flexible RRAM-based computation. A passive 54x108 RRAM crossbar array performs VMM in the analog domain. Specialized mixed-signal circuits stimulate and read the outputs of the RRAM crossbar. The single-chip CMOS prototype includes a RISC processor interfaced to a memory-mapped mixed-signal core. In the mixed-signal core, ADCs and DACs interface to the passive RRAM crossbar. The RISC processor controls the mixed-signal circuits and the algorithm data path. The system is fully programmable and supports forward and backward propagation. As proof of concept, a fully-integrated 0.18μm CMOS prototype with a post-processed RRAM array demonstrates several key functions of machine learning including online learning. The mixed-signal core consumes 64mW at an operating frequency of 148MHz. The total system power consumption, considering the mixed-signal circuitry, the digital processor, and the passive RRAM array is 307mW. The maximum theoretical throughput is 2.6 GOPS at an efficiency of 8.5 GOPS/W.
- Top Cited Papers
- 14 October 2019
- Published by Springer Nature in Nature Electronics
- Vol. 2 (10) , 480-487
Temporal data classification and forecasting using a memristor-based reservoir computing system
- 14 October 2019
- Published by Springer Nature in Nature Electronics
- Vol. 2 (10) , 480-487
Top Cited PapersTemporal data classification and forecasting using a memristor-based reservoir computing system
- Top Cited Papers
- 15 July 2019
- Published by Springer Nature in Nature Electronics
- Vol. 2 (7) , 290-299
A fully integrated reprogrammable memristor–CMOS system for efficient multiply–accumulate operations
- 15 July 2019
- Published by Springer Nature in Nature Electronics
- Vol. 2 (7) , 290-299
Top Cited PapersA fully integrated reprogrammable memristor–CMOS system for efficient multiply–accumulate operations
- 1 January 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
Feature extraction and analysis using memristor networks
Approaches toward feature extraction and image analysis using memristor networks will be discussed. Through hardware implementation of a sparse coding algorithm in a fabricated 32×32 memristor array, lateral inhibition among neurons is obtained and allows the network to settle to a more optimal, sparse solution from many possible solutions. This capability enables the network to identify the hidden features that constitute the input real-world image. By utilizing the internal dynamics of the device, temporal features can also be learned and processed.
- 1 January 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
Feature extraction and analysis using memristor networks
Approaches toward feature extraction and image analysis using memristor networks will be discussed. Through hardware implementation of a sparse coding algorithm in a fabricated 32×32 memristor array, lateral inhibition among neurons is obtained and allows the network to settle to a more optimal, sparse solution from many possible solutions. This capability enables the network to identify the hidden features that constitute the input real-world image. By utilizing the internal dynamics of the device, temporal features can also be learned and processed.
- Top Cited PapersOpen Access
- 19 December 2017
- Published by Springer Nature in Nature Communications
- Vol. 8 (1) , 1-10
Reservoir computing using dynamic memristors for temporal information processing
Reservoir computing systems utilize dynamic reservoirs having short-term memory to project features from the temporal inputs into a high-dimensional feature space. A readout function layer can then effectively analyze the projected features for tasks, such as classification and time-series analysis. The system can efficiently compute complex and temporal data with low-training cost, since only the readout function needs to be trained. Here we experimentally implement a reservoir computing system using a dynamic memristor array. We show that the internal ionic dynamic processes of memristors allow the memristor-based reservoir to directly process information in the temporal domain, and demonstrate that even a small hardware system with only 88 memristors can already be used for tasks, such as handwritten digit recognition. The system is also used to experimentally solve a second-order nonlinear task, and can successfully predict the expected output without knowing the form of the original dynamic transfer function.
- 19 December 2017
- Published by Springer Nature in Nature Communications
- Vol. 8 (1) , 1-10
Top Cited PapersOpen AccessReservoir computing using dynamic memristors for temporal information processing
Reservoir computing systems utilize dynamic reservoirs having short-term memory to project features from the temporal inputs into a high-dimensional feature space. A readout function layer can then effectively analyze the projected features for tasks, such as classification and time-series analysis. The system can efficiently compute complex and temporal data with low-training cost, since only the readout function needs to be trained. Here we experimentally implement a reservoir computing system using a dynamic memristor array. We show that the internal ionic dynamic processes of memristors allow the memristor-based reservoir to directly process information in the temporal domain, and demonstrate that even a small hardware system with only 88 memristors can already be used for tasks, such as handwritten digit recognition. The system is also used to experimentally solve a second-order nonlinear task, and can successfully predict the expected output without knowing the form of the original dynamic transfer function.
- 1 July 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 19-20
Epsilon-greedy strategy for online dictionary learning with realistic memristor array constraints
The influence of realistic experimental constraints of memristor crossbar array on dictionary learning was investigated. A solution with epsilon-greedy strategy was proposed to improve training.
- 1 July 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 19-20
Epsilon-greedy strategy for online dictionary learning with realistic memristor array constraints
The influence of realistic experimental constraints of memristor crossbar array on dictionary learning was investigated. A solution with epsilon-greedy strategy was proposed to improve training.