17 November 2022 in IEEE Journal on Emerging and Selected Topics in Circuits and Systems
Memristors show great potential for being integrated into CMOS technology and provide new approaches for designing computing-in-memory (CIM) systems, brain-inspired applications, trimming circuits and other topologies for the beyond-CMOS era. A crucial characteristic of the memristor is multi-state 1 switching. Memristors are capable of representing information in an ultra-compact fashion, by storing multiple bits per device. However, certain challenges remain in multi-state memristive circuits and systems design such as device stability and peripheral circuit complexity. In this paper, we review the state of the art of multi-state memristor technologies and their associated CMOS/Memristor circuit design, and discuss the challenges regarding device imperfection factors, modelling, peripheral circuit design and layout. We present measurement results of our in-house fabricated multi-state memristor as an example to further illustrate the feasibility of applying multi-state memristors in CMOS design, and demonstrate their related future applications such as multi-state memristive memories in machine learning, memristive neuromorphic applications, trimming and tuning circuits, etc. In the end, we summarize past and present efforts done in this field and envisage the direction of multi-state memristor related research.
4 September 2024 in IEEE Transactions on Circuits and Systems I: Regular Papers
Content addressable memory is popular in intelligent computing systems as it allows parallel content-searching in memory. Emerging CAMs show a promising increase in bitcell density and a decrease in power consumption than pure CMOS solutions. This article introduced an energy-efficient 3T1R1C TCAM cooperating with capacitor dividers and RRAM devices. The RRAM as a storage element also acts as a switch to the capacitor divider while searching for content. CAM cells benefit from working parallel in an array structure. We implemented a $64\times 64$ array and digital controllers to perform with an internal built-in clock frequency of 875MHz. Both data searches and reads take three clock cycles. Its worst average energy for data match is reported to be 1.71fJ/bit-search and the worst average energy for data miss is found at 4.69fJ/bit-search. The prototype is simulated and fabricated in 0.18um technology with in-lab RRAM post-processing. Such memory explores the charge domain searching mechanism and can be applied to data centers that are power-hungry.
8 July 2025 in IEEE Open Journal of Circuits and Systems
Spike sorting is a critical process for decoding large-scale neural activity from extracellular recordings. The advancement of neural probes facilitates the recording of a high number of neurons with an increase in channel counts, arising a higher data volume and challenging the current on-chip spike sorters. This paper introduces L-Sort, a novel on-chip spike sorting solution featuring median-of-median spike detection and localization-based clustering. By combining the median-of-median approximation and the proposed incremental median calculation scheme, our detection module achieves a reduction in memory consumption. Moreover, the localization-based clustering utilizes geometric features instead of morphological features, thus eliminating the memory-consuming buffer for containing the spike waveform during feature extraction. Evaluation using Neuropixels datasets demonstrates that L-Sort achieves competitive sorting accuracy with reduced hardware resource consumption. Implementations on FPGA and ASIC (180 nm technology) demonstrate significant improvements in area and power efficiency compared to state-of-the-art designs while maintaining comparable accuracy. If normalized to 22 nm technology, our design can achieve roughly ×10 area and power efficiency with similar accuracy, compared with the state-of-the-art design evaluated with the same dataset. Therefore, L-Sort is a promising solution for real-time, high-channel-count neural processing in implantable devices.
2 April 2026 in IEEE Journal on Exploratory Solid-State Computational Devices and Circuits
Artificial Intelligence models are currently driven by a significant up-scaling of their complexity, with massive matrix-multiplication workloads representing the major computational bottleneck. In-memory computing architectures are proposed to avoid the Von Neumann bottleneck. However, both digital/binary-based and analogue in-memory computing architectures suffer from various limitations, which significantly degrade the performance and energy efficiency gains. This work proposes OISMA, an energy-efficient in-memory computing architecture that utilizes the computational simplicity of a quasi-stochastic computing domain (Bent-Pyramid system), while keeping the same efficiency, scalability, and productivity of digital memories. OISMA converts normal memory read operations into in-situ stochastic multiplication operations with a negligible cost. An accumulation periphery then accumulates the output multiplication bitstreams, achieving the matrix multiplication functionality. A 4KB 1T1R OISMA array was implemented using a commercial 180nm technology node and in-house RRAM technology. At 50 MHz, it achieves 0.789 TOPS/W and 3.98 GOPS/mm2 for energy and area efficiency, respectively, occupying an effective computing area of 0.804241 mm2. Scaling OISMA to 22nm technology shows a significant improvement of two orders of magnitude in energy efficiency and one order of magnitude in area efficiency, compared to dense matrix multiplication in-memory computing architectures.
28 April 2026 in IEEE Transactions on Circuits and Systems I: Regular Papers
Emerging resistive switching devices require large-scale characterization, yet existing approaches often rely on external instrumentation, small custom arrays or architectures that lack autonomous current sensing. This work introduces MELITA, a fully integrated CMOS platform that supports parallel characterization of more than one million two-terminal devices. Fabricated in $\mathrm {180~ nm }$ CMOS, the chip combines on-chip DACs, current-to-voltage converters, column-parallel ADCs, and a digital controller into a unified readout architecture. Automated current ranging extends measurable input currents across five decades, while programmable read voltages enable flexible test conditions. Measured results confirm both block-level and system-level operation, with a 12 bit digital output. Application-specific experiments demonstrate the ability to extract I-V curves directly from large-scale two-terminal device arrays composed of passive top-metal structures, validating scalability and readout functionality of the platform. Although noise limits the effective Signal-to-Noise Ratio (SNR) to 43dB, the platform uniquely integrates array access, analog front-end, digitization, and control into a scalable test vehicle. MELITA therefore provides a practical step toward realistic benchmarking of resistive devices within system-level CMOS environments.