XQXin Qiao

Xin Qiao

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Xin Qiao has a total of 31 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • MixCIM: A Hybrid Computing-in-Memory Macro With Less Data-Movement and Better Memory-Reuse for Depthwise Separable Neural Networks

    by Xin Qiao, Youming Yang, Jiahao Song, Renjie Wei, Xiyuan Tang, Meng Li, Runsheng Wang, Yuan Wang

    Computing-in-memory (CIM) architectures have demonstrated strong potential for edge artificial intelligence (AI) devices due to their enhanced parallelism and energy efficiency. With the growing complexity of AI tasks and the rapid increase in model size, computation and deployment costs have surged. Depthwise separable neural networks (DSNNs) have attracted interest for offering appreciable performance with far fewer parameters. However, the core operation in DSNNs, depthwise separable convolution (DSC), exhibits a different multiply-and-accumulate (MAC) pattern that conventional CIM macros are not optimized to handle efficiently. As a result, deploying DSC on conventional CIM platforms remains challenging. In this article, we propose MixCIM, a mixed macro that integrates computing-near-memory (CNM) and CIM. It achieves high energy efficiency for DSC computations through a mix-memory computing architecture. With a 2’s complement storage-sign magnitude computation (TCS-SMC) encoding, it reduces the activations’ movement and MAC energy. A slidable eDRAM array enhances the parallelism of DW convolution to improve throughput. A memory sharing scheme between input and output activations increases effective memory use. It optimizes complete DSC with both PW and DW convolutions, achieving superior energy efficiency. Measurement results show that the proposed 28-nm macro achieves 40.99 TOPS/W for pointwise convolution (PWC), 15.31 TOPS/W for depthwise convolution (DWC), and 17.2 TOPS/W for DSC, all at signed 8-bit inputs and weights. The accuracy on CIFAR-10 is 92.47% for MobileNet-V2.


  • An eDRAM-Based Computing-in-Memory Macro With Full-Valid-Storage and Channel-Wise-Parallelism for Depthwise Neural Network

    by Xin Qiao, Youming Yang, Chang Xue, Yandong He, Xiaoxin Cui, Song Jia, Yuan Wang

    Computing-in-memory (CIM) provides a highly efficient solution for neural networks in edge artificial intelligence applications. Most SRAM-based CIM designs can achieve high energy efficiency and area efficiency in standard convolutional layer’s multiply-and-accumulate (MAC) operations. However, when deploying depthwise separable convolution, they face several challenges. For these CIMs with weight-stationary, the lower activations reuse increases redundant memory reducing area efficiency and the fewer parameters of depthwise convolutional MAC decreases energy efficiency. To address these issues, we propose a depthwise separable convolutional computing-in-memory (DSC-CIM) that supports channel-wise parallel computation to increase area efficiency and energy efficiency. It includes three key techniques: (1) a 5T2C eDRAM bitcell for low power activation update and high area efficiency, (2) an independent update in the column direction to enable horizontal and vertical movement of the convolution window in the feature map, and (3) a data weight configuration circuit (DWCC) that supports both signed and unsigned parameters’ MAC operation. Layout post-simulations show that the proposed 28 nm DSC-CIM macro achieves an energy efficiency of 20.13 TOPS/W for 8b parameters on depthwise convolution. The inference accuracy on CIFAR-10 with 8b MobileNet-V2 model is 92.6%.


  • A 28nm 8Kb Reconfigurable SRAM Computing-In-Memory Macro With Input-Sparsity Optimized DTC for Multi-Mode MAC Operations

    by Kanglin Xiao, Xin Qiao, Xiaoxin Cui, Jiahao Song, Haoyang Luo, Xin’aN Wang, Yuan Wang

    Computing-in-memory (CIM) is an emerging approach for alleviating von Neumann bottleneck and improving energy efficiency and throughput. This brief presents a reconfigurable SRAM CIM macro supporting multi-mode multiply-and-accumulate (MAC) operations, including binary weight network (BWN) MAC, ternary weight network (TWN) MAC, and multi-bit MAC operations. The BWN and the TWN topology are proposed for signed 5b-input MAC operation. The multi-bit MAC operation is achieved by both positive and inverse code with AND operation, which improves hardware utilization and doubles the throughput. To overcome the different dynamic ranges caused by reconfigurations, we proposed two adaptive solutions for different discharging capacitors, and an input-sparsity digital-to-time converter with intrinsic delay elimination to improve energy efficiency and linearity. The proposed 8Kb macro is fulfilled in a 28 nm CMOS with an energy efficiency of 1773.5 TOPS/W in BWN mode. The test chip achieves an accuracy of 84.35% on the CIFAR-10 dataset at 4b precision in inputs and weights.


  • A 16.38TOPS and 4.55POPS/W SRAM Computing-in-Memory Macro for Signed Operands Computation and Batch Normalization Implementation

    by Xin Qiao, Qingyu Guo, Xiyuan Tang, Jiahao Song, Renjie Wei, Meng Li, Runsheng Wang, Yuan Wang

    Edge artificial intelligence applications impose rigorous demands on local hardware to improve throughput and energy efficiency. Computing-in-memory (CIM) architectures provide high parallel and energy-efficient solutions to accelerate the multiply-and-accumulate (MAC) operations in neural networks (NNs). While SRAM-based charge-domain CIM is achieving thousands of TOPS/W energy efficiency, it encounters limitations when dealing with full NN model deployments where both activations and weights are signed. This paper proposes an SRAM-based signed batch normalization (BN) CIM macro for supporting efficient bitwise sparse MAC computation with signed operands and BN operations in deep neural networks. The key features of this macro encompass: 1) a multibit weight unit for the optimization of bitstream sparsity and the sign bit computation, 2) a 2b-serial input configuration to increase throughput and the ADC energy amortization, and 3) a quantization-hardware co-design for the BN implementation. Measurement results show that the proposed 28 nm 64 Kb CIM macro achieves 16.38 TOPS throughput and 4.55 POPS/W energy efficiency, both normalized to 1b operands. The test accuracy of CIFAR10 is 92%, based on the ResNet18 model with co-design BN implementation at signed-8b precision activations and weights.


  • A 768.7-2124.2 TOPS/W Time-Domain Computing-in-Memory Macro With Low Static Leakage and Precision-Configurable TDC

    by Chang Xue, Xin Qiao, Xu Ren, Gang Du, Yuan Wang, Yandong He

    In always-on edge devices, the static leakage current in weights retention accounts for a significant proportion of power consumption. Furthermore, the analog-to-digital converter (ADC) module consumes a lot of area and power in most computing-in-memory (CIM) architectures. Therefore, this design presents an 8 Kb time-domain (TD) CIM macro with low static leakage. In addition, this work takes advantage of the scalability of TD-CIM and presents a precision-configurable time-to-digital converter (TDC) applied to scenarios that require adjustable accuracy. It achieves configurable precision of 2 to 4 bits with a multiplication-and-accumulation (MAC) range of 32 in each delay chain. The TDC module accounts for only 2.9% of the total macro area, achieving nearly a tenfold improvement over prior works. The measured normalized energy efficiency is 768.7-2124.2 TOPS/W. The static leakage of the macro is 206 fW/bit while VDD is 1.2 V. This work achieves 91.1%-86.1% inference accuracy on the CIFAR-10 dataset with 8-bit precision in inputs and weights.


  • A 4-bit Calibration-Free Computing-In-Memory Macro With 3T1C Current-Programed Dynamic-Cascode Multi-Level-Cell eDRAM

    by Jiahao Song, Xiyuan Tang, Haoyang Luo, Haoyi Zhang, Xin Qiao, Zixuan Sun, Xiangxing Yang, Zihan Wu, Yuan Wang, Runsheng Wang, et al.

    Analog computing-in-memory (CIM) has been widely explored for computing neural networks (NNs) efficiently. However, most analog CIM implementations trade compute accuracy for energy efficiency. The low accuracy restricts the practical application of analog CIM. In this article, a current-programming CIM that unifies the weight programming and computing in the current domain is proposed to address this dilemma. The enabled technique is a novel 3-transistor 1-capacitor (3T1C) embedded dynamic random access memory (eDRAM) cell. The current-programming mechanism and the dynamic-cascode read structure of the 3T1C cell make it immune to transistor-level non-idealities, including nonlinear $I$ – $V$ , threshold voltage variations, and short-channel effect. Therefore, the cell enables multi-level-cell (MLC) operations without any calibration, supporting eight current-weight levels (0–700 nA). In addition, a voltage–current two-step programming scheme is proposed to boost the sub-microamphere current-weight writing speed. To support signed 4-b weights, a pseudo-differential CIM cell composed of two 3T1C MLCs is developed. Fabricated in a 65-nm CMOS, the prototype demonstrates 2.2 $\times $ reduction in macro-level variation through current programming. Benefiting from sub-microamphere compute currents, the prototype achieves the 4-b energy efficiencies of 233–304 TOPS/W. With a refresh interval of 0.4 ms, the macro achieves >90% inference accuracy on CIFAR10.


  • OASIS: A 28-nm 32-kb SRAM-Based Computing-in-Memory Design With Output Activation Sparsity Support

    by Qingyu Guo, Nanbing Pan, Xin Qiao, Xiaoxin Cui, Yuan Wang

    Computing-in-memory (CIM) has emerged as a promising architecture for energy-efficient edge neural network inference. While CIM designs have shown great potential to benefit from sparsity, there are still limitations in supporting output analog-digital conversion (ADC) sparsity in current designs due to redundant circuits and weight-refreshing overhead. In this brief, we propose a novel CIM design called OASIS (Output Activation SparsIty Support). OASIS employs a single-bit-line computation circuit to support output ADC sparsity, avoiding redundancy or overhead. Our design achieves up to 812.2 TOPS/W normalized energy efficiency at a typical 50% sparsity. We demonstrate its effectiveness by deploying all layers of quantized neural networks to the chip and evaluating them on the MNIST and CIFAR-10 datasets, achieving 99.20% and 84.33% accuracy, respectively.


  • A 28 nm 16 Kb Bit-Scalable Charge-Domain Transpose 6T SRAM In-Memory Computing Macro

    by Jiahao Song, Xiyuan Tang, Xin Qiao, Yuan Wang, Runsheng Wang, Ru Huang

    This article presents a compact, robust, and transposable SRAM in-memory computing (IMC) macro to support feed forward (FF) and back propagation (BP) computation within a single macro. The transpose macro is created with a clustering structure, and eight 6T bitcells are shared with one charge-domain computing unit (CCU) to efficiently deploy the DNNs weights. The normalized area overhead of clustering structure compared to 6T SRAM cell is only 0.37. During computation, the CCU performs robust charge-domain operations on the parasitic capacitances of the local bitlines in the IMC cluster. In the FF mode, the proposed design supports 128-input 1b XNOR and 1b AND multiplications and accumulations (MACs). The 1b AND can be extended to multi-bit MAC via bit-serial (BS) mapping, which can support DNNs with various precision. A power-gated auto-zero Flash analog-to-digital converter (ADC) reducing the input offset voltage maintains the overall energy efficiency and throughput. The proposed macro is prototyped in a 28-nm CMOS process. It demonstrates a 1b energy efficiency of $166\vert 257$ TOPS/W in FF-XNOR $\vert $ AND mode, and 31.8 TOPS/W in BP mode, respectively. The macro achieves $80.26\% \vert 85.07\%$ classification accuracy for the CIFAR-10 dataset with 1b $\vert 4\text{b}$ CNN models. Besides, 95.50% MNIST dataset classification accuracy (95.66% software accuracy) is achieved by the BP mode of the proposed transpose IMC macro.


  • A 28nm 32Kb SRAM Computing-in-Memory Macro With Hierarchical Capacity Attenuator and Input Sparsity-Optimized ADC for 4b Mac Operation

    by Kanglin Xiao, Xiaoxin Cui, Xin Qiao, Jiahao Song, Haoyang Luo, Xin’an Wang, Yuan Wang

    Computing-in-memory (CIM) is an emerging approach for alleviating the Von-Neumann bottleneck of latency and energy overheads, and improving energy efficiency and throughput. In this brief, we present a novel CIM macro aimed at improving the energy efficiency and throughput of edge devices when running 4b multiply-and-accumulate (MAC) operations. The proposed architecture uses (1) a customized 9T1C bit-cell in charge-domain computation for sensing margin improvement and compact design; (2) a hierarchical capacity attenuator for 4b weight accumulation without complicated controlling switches and signals for throughput improvement; (3) an input sparsity-sensing-based flash analog-to-digital converters readout scheme to improve energy efficiency and throughput. Fabricated in 28nm CMOS technology, the proposed 32Kb SRAM CIM macro demonstrates an average energy efficiency of 646.6 TOPS/W (normalized to 4b/4b input/weight) and a throughput of 1638.4 GOPS while achieving 84.89% classification accuracy on the CIFAR-10 dataset at 4b precision in inputs and weights.