XLXinyi Li

Xinyi Li

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

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Xinyi Li has a total of 19 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • NMIX: NoC-Aware Mixed Precision Quantization for Energy-Efficient ML Accelerator

    by Xinyi Li, Eunbin Park, Myeongjun Lee, Wenjie Fan, Yuxiang Fu, Li Li, Youngjoo Lee

    Energy efficiency remains a critical challenge in network-on-chip (NoC) based machine learning (ML) accelerators due to high computation and communication costs. Existing NoC optimization techniques often overlook the intrinsic co-optimization between NoC characteristics and the ML model design. We propose NoC-aware mixed precision quantization (NMIX), a two-phase algorithm that jointly minimizes computation and routing energy while preserving model accuracy. The first phase performs a coarse search to identify accuracy-sensitive quantization boundaries, followed by fine-grained precision assignment based on sensitivity and NoC energy profiles. Experiments on image classification tasks with VGG and ResNet demonstrate energy savings of up to 91.5% over FP32 and 57.5% over uniform 8-bit quantization, confirming the scalability and effectiveness of NMIX for energy-efficient ML accelerators.

  • Open Access

    Compact Interleaved Thermal Control for Improving Throughput and Reliability of Networks-on-Chip

    by Tong Cheng, Zirui Xu, Xinyi Li, Li Li, Yuxiang Fu

    Due to the scaling of sub-micron technology and the growing complexity of applications, escalating power density and traffic workloads heavily burden the network-on-chip (NoC) in multi-core systems and exacerbate thermal reliability issues. While recent thermal management techniques offer innovative solutions, they often employ the same management strategy for all tiles in NoC and activate it synchronously, which inevitably causes system oscillation and temperature cycling. In this paper, we propose a novel compact interleaved thermal control method that staggers the control phases of neighboring nodes to create negative feedback for each tile. We further explore the optimal control phase assignment by formulating it as a graph coloring problem to achieve the best performance. Experimental results demonstrate that the proposed method lowers the maximal spatial and temporal temperature variations up to 83.1% and 71.2% and improves the system throughput by 35.85% averagely compared with the state-of-the-art work. Besides, the proposed method can achieve a significant average enhancement of 317.50% and 234.83% in the minimal and average thermal-related mean time to failure (MTTF). Moreover, the method is scalable without extra power or area costs and is compatible with existing thermal management techniques.

  • Automatic Generation and Optimization Framework of NoC-Based Neural Network Accelerator Through Reinforcement Learning

    by Yongqi Xue, Jinlun Ji, Xinming Yu, Shize Zhou, Siyue Li, Xinyi Li, Tong Cheng, Shiping Li, Kai Chen, Zhonghai Lu, et al.

    Choices of dataflows, which are known as intra-core neural network (NN) computation loop nest scheduling and inter-core hardware mapping strategies, play a critical role in the performance and energy efficiency of NoC-based neural network accelerators. Confronted with an enormous dataflow exploration space, this paper proposes an automatic framework for generating and optimizing the full-layer-mappings based on two reinforcement learning algorithms including A2C and PPO. Combining soft and hard constraints, this work transforms the mapping configuration into a sequential decision problem and aims to explore the performance and energy efficient hardware mapping for NoC systems. We evaluate the performance of the proposed framework on 10 experimental neural networks. The results show that compared with the direct-X mapping, the direct-Y mapping, GA-base mapping, and NN-aware mapping, our optimization framework reduces the average execution time of 10 experimental NNs by 9.09%, improves the throughput by 11.27%, reduces the energy by 12.62%, and reduces the time-energy-product (TEP) by 14.49%. The results also show that the performance enhancement is related to the coefficient of variation of the neural network to be computed.


  • Open Access

    TTNNM: Thermal- and Traffic-Aware Neural Network Mapping on 3D-NoC-based Accelerator

    by Xinyi Li, Wenjie Fan, Heng Zhang, Jinlun Ji, Tong Cheng, Shiping Li, Li Li, Yuxiangfu Fu

    3D Network on Chips (3D-NoCs) have ample on-chip wiring resources and high bandwidth, yet face numerous hotspots and higher temperature gradients due to increased integration and power density. This could lead to device failure, impacting system stability. Our paper introduces a thermal- and traffic-aware mapping method for 3D-NoC-based neural network accelerators. Firstly, based on the average load of different neural network layer, we determine their mapping sequences and suitable dies. Secondly, to minimize delay and alleviate hotspot temperatures, we allocate groups to appropriate nodes. Compared with previous works, TTNNM reduces the average temperature by 3.0°C, 2.2°C, 2.4°C, temperature variance by 58.4%, 64.8%, 73.0%, maximum temperature by 9.3°C, 7.9°C, 12.0°C, and packet latency by 31.7%, 17.2%, 25.1%.


  • A Nearest-Neighbor-Based Thermal Sensor Allocation and Temperature Reconstruction Method for 3-D NoC-Based Multicore Systems

    by Menghao Guo, Tong Cheng, Xinyi Li, Li Li, Yuxiang Fu

    To avoid overheating of three-dimensional Network-on-Chip (NoC)-based multicore systems, many researchers have used Dynamic Thermal Management (DTM) techniques, which need embedded thermal sensors to provide accurate temperature information. However, only a few sensors can be embedded due to the limited hardware cost. So, it is crucial to find an appropriate way to allocate number-limited sensors at design time and reconstruct full-chip temperature accurately using the limited temperature information. However, the relationship between the non-sensor-allocated nodes and the sensor-allocated nodes modeled by the existing methods has a deviation from the actuality, which leads to an inaccurate temperature reconstruction. Another problem is that the existing methods depend highly on the training data. The estimation error can be significant when the running application’s traffic characteristic differs from the one in the offline phase. This paper presents a sensor allocation method based on the cores’ spatial correlation, which is not dependent on the training data. Our allocation method contains two stages: 1) Using our Nearest-Neighbor-based initialization algorithm to allocate sensors preliminarily; 2) Using Genetic Algorithm (GA) to optimize the initial allocation. Besides, we use Artificial Neural Network (ANN) to reconstruct the full-chip temperature. Compared with the state-of-the-art methods, our method can improve the average accuracy of the estimated temperature under different scenarios by 17.60% to 88.63%. What’s more, our approach has high flexibility and can adapt to different application scenarios with high accuracy with only one offline training.