MWMaolin Wang

Maolin Wang

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

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Maolin Wang has a total of 41 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    GraFlex: Flexible Graph Processing on FPGAs through Customized Scalable Interconnection Network

    by Chunyou Su, Linfeng Du, Tingyuan Liang, Zhe Lin, Maolin Wang, Sharad Sinha, Wei Zhang

    Graph processing system design has been widely considered to be a challenging topic due to the mismatch between the computational throughput requirement and the memory bandwidth. Recent works try to deliver better graph processing systems by taking advantage of application-specific architectures and emerging high-bandwidth memory on FPGAs. However, there is still ample room for improvements regarding flexibility, scalability, and usability. This paper presents GraFlex, a flexible scatter-gather graph processing framework on FPGAs with scalable interconnection networks. It adopts the Bulk-Synchronous Parallel (BSP) paradigm for global control and synchronization, enabling rapid deployment of performant graph processing systems through HLS-based design flows. GraFlex conducts software-hardware co-optimization to boost system performance. It configures the compact graph format, partition scheme, and memory channel allocation strategy to support scalable designs. Resource-efficient multi-stage butterfly interconnection network achieves on-device data communication and facilitates throughput matching. To handle fragmented memory requests, we propose coalesced memory access engines to improve bandwidth utilization. GraFlex is comprehensively evaluated with various graph applications and real-world datasets. Our results show up to 2.09\texttimes average speedup in traversal throughput over the existing state-of-the-art work with a non-negligible reduction in power and resource consumption. A case study of the breadth-first search (BFS) application shows a 6.58\texttimes speedup in average algorithm throughout with proper implementation choices enabled by the scatter-gather mechanism implemented. The BFS study also reports an almost linear throughput scaling versus the number of processing elements (PEs) and memory channels.


  • Quality Prediction of Rocket Engine Combustion Chamber Shell Spinning Process Based on CGAN and Hybrid Neural Network

    by Junbo Yi, Xin He, Maolin Wang, Xiujuan Zheng, Feng Guo, Yong Zhang

    The combustion chamber shell is one of the main structures of rocket engines, and the quality prediction of its spinning process is crucial to improve the stability of the engine and the subsequent improvement process. In this article, we propose a quality prediction method based on data augmentation and a hybrid neural network model combining feature attention (FA) and temporal convolutional network and long short term memory network (TCN-LSTM). First, the feature multi-collinearity problem is solved by principal component analysis dimensionality reduction of the equipment time series data of the shell machining process. Then, conditional generative adversarial network is used for data augmentation to generate more samples using quality indicators as condition information to solve the problem of insufficient data collection in actual engineering. Finally, the key feature information is enhanced using the FA mechanism, while the dual quality metrics are predicted by TCN-LSTM, and the prediction results are corrected using the least squares method. In this paper, a real rocket engine combustion chamber shell machining dataset is used to validate the model effect, and the experimental results show that the method has higher prediction accuracy by comparing with different machine learning methods.

  • Few-Shot Gearbox Fault Diagnosis Based on Semi-Supervised Prototypical Network and Contrastive Learning

    by Shulie Cheng, Yingjie Liang, Zuowei Ping, Xuechao Liao, Maolin Wang, Yong Zhang

    Due to the scarcity of labeled faulty data in industrial practice, fault diagnosis models often face challenges related to overfitting and limited accuracy. This article introduces a novel solution to tackle the problem—a fault diagnosis method that leverages a combination of a semi-supervised prototype network and contrastive learning. Firstly, a limited number of labeled samples is employed to construct pairs of positive and negative samples. And the training method of contrastive learning is used to provide appropriate initial parameters for the autoencoder. Secondly, the autoencoder is employed as a feature mapping function within the prototype network and obtains the prototype using a limited number of labeled samples. Finally, a prototype refinement method fine-tunes the prototype through unlabeled data, reducing the impact of anomalous data and obtaining a more stable and accurate prototype. The effectiveness of the proposed method is evaluated using gearbox data as a validation dataset. Through comparative analysis with other methods, the results demonstrate that the proposed approach achieves superior diagnostic accuracy.

  • Atomically Dispersed N/O-Coordinated Cobalt Catalyst Enables Aerobic Oxygenation of Olefins under Ambient Conditions

    by Kaiyue Zhao, Qikai Shen, Yu Tao, Jieran Li, Maolin Wang, Chen Li, Bingjun Xu


  • Open Access

    RSQP: Problem-specific Architectural Customization for Accelerated Convex Quadratic Optimization

    by Maolin Wang, Ian McInerney, Bartolomeo Stellato, Stephen Boyd, Hayden Kwok-Hay So

    Convex optimization is at the heart of many performance-critical applications across a wide range of domains. Although many high-performance hardware accelerators have been developed for specific optimization problems in the past, designing such accelerator is a challenging task and the resulting computing architecture is often so specific to the targeted application that they can hardly be reused even in a related application within the same domain. To accelerate general-purpose optimization solvers that must operate on diverse user input during run time, an ideal hardware solver should be able to adapt to the provided optimization problem dynamically while achieving high performance and power-efficiency. In this work, a hardware-accelerated general-purpose quadratic program solver, called RSQP, with reconfigurable functional units and data path that facilitate problem-specific customization is presented. RSQP uses a string-based encoding to describe the problem structure with fine granularity. Based on this encoding, functional units and datapath customized to the sparsity pattern of the problem are created by solving a dictionary-based lossless string compression problem and a mixed integer linear program respectively. RSQP has been integrated to accelerate the general-purpose quadratic programming solver OSQP and has been tested using an extensive benchmark with 120 optimization problems from 6 application domains. Through architectural customization, RSQP achieves up to 7× performance improvement over its baseline generic design. Furthermore, when compared with a CPU and a GPU-accelerated implementation, RSQP achieves up to 31.2× and 6.9× end-to-end speedup on these benchmark programs, respectively. Finally, the FPGA accelerator operates at up to 6.6× lower dynamic power consumption and up to 22.7× higher power efficiency over the GPU implementation, making it an attractive solution for power-conscious datacenter applications.


  • NITI: Training Integer Neural Networks Using Integer-Only Arithmetic

    by Maolin Wang, Seyedramin Rasoulinezhad, Philip H. W. Leong, Hayden K.-H. So

    Low bitwidth integer arithmetic has been widely adopted in hardware implementations of deep neural network inference applications. However, despite the promised energy-efficiency improvements demanding edge applications, the use of low bitwidth integer arithmetic for neural network training remains limited. Unlike inference, training demands high dynamic range and numerical accuracy for high quality results, making the use of low-bitwidth integer arithmetic particularly challenging. To address this challenge, we present a novel neural network training framework called NITI that exclusively utilizes low bitwidth integer arithmetic. NITI stores all parameters and accumulates intermediate values as 8-bit integers while using no more than 5 bits for gradients. To provide the necessary dynamic range during the training process, a per-layer block scaling exponentiation scheme is utilized. By deeply integrating with the rounding procedures and integer entropy loss calculation, the proposed scaling scheme incurs only minimal overhead in terms of storage and additional computation. Furthermore, a hardware-efficient pseudo-stochastic rounding scheme that eliminates the need for external random number generation is proposed to facilitate conversion from wider intermediate arithmetic results to lower precision for storage. Since NITI operates only with standard 8-bit integer arithmetic and storage, it is possible to accelerate it using existing low bitwidth operators originally developed for inference in commodity accelerators. To demonstrate this, an open-source software implementation of end-to-end training, using native 8-bit integer operations in modern GPUs is presented. In addition, experiments have been conducted on an FPGA-based training accelerator to evaluate the hardware advantage of NITI. When compared with an equivalent training setup implemented with floating point storage and arithmetic, NITI has no accuracy degradation on the MNIST and CIFAR10 datasets. On ImageNet, NITI achieves similar accuracy as state-of-the-art integer training frameworks without relying on full-precision floating-point first and last layers.


  • Open Access

    Low-Latency In Situ Image Analytics With FPGA-Based Quantized Convolutional Neural Network

    by Maolin Wang, Kelvin C. M. Lee, Bob M. F. Chung, Sharatchandra Varma Bogaraju, Ho-Cheung Ng, Justin S. J. Wong, Ho Cheung Shum, Kevin K. Tsia, Hayden Kwok-Hay So

    Real-time in situ image analytics impose stringent latency requirements on intelligent neural network inference operations. While conventional software-based implementations on the graphic processing unit (GPU)-accelerated platforms are flexible and have achieved very high inference throughput, they are not suitable for latency-sensitive applications where real-time feedback is needed. Here, we demonstrate that high-performance reconfigurable computing platforms based on field-programmable gate array (FPGA) processing can successfully bridge the gap between low-level hardware processing and high-level intelligent image analytics algorithm deployment within a unified system. The proposed design performs inference operations on a stream of individual images as they are produced and has a deeply pipelined hardware design that allows all layers of a quantized convolutional neural network (QCNN) to compute concurrently with partial image inputs. Using the case of label-free classification of human peripheral blood mononuclear cell (PBMC) subtypes as a proof-of-concept illustration, our system achieves an ultralow classification latency of 34.2 $\mu \text{s}$ with over 95% end-to-end accuracy by using a QCNN, while the cells are imaged at throughput exceeding 29 200 cells/s. Our QCNN design is modular and is readily adaptable to other QCNNs with different latency and resource requirements.


  • An integrated method based on relevance vector machine for short-term load forecasting

    by Jia Ding, Maolin Wang, Zuowei Ping, Dongfei Fu, Vassilios S. Vassiliadis