ILIng-Chao Lin

Ing-Chao Lin

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

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Ing-Chao Lin has a total of 65 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • TT-RRAM: Joint Improvement of Sparsity and Repetition in Tensor-Train Inference With Fused Processing on RRAM Accelerators

    by Fang-Yi Gu, Pin-Hong Liu, Ing-Chao Lin, Bo Yuan

    Deep neural networks compressed using tensor-train decomposition, known as TT-DNNs, significantly reduce model size to decrease storage requirements but suffer from more frequent data movement during inference. To address this issue, resistive random access memory (RRAM)-based deep neural network (DNN) accelerators, by leveraging computing-in-memory, can mitigate the data movement and efficiently execute the vector–matrix multiplication (VMM) operations required during inference, making them an ideal candidate for accelerating the inference for TT-DNNs. However, RRAM-based accelerators face three challenges in TT-format DNN inference: 1) weights distribution on the RRAM crossbar with low sparsity and repetition; 2) low input sparsity; and 3) high data dependencies and storage overhead during inference. To address these challenges, we propose three corresponding methods: 1) base-offset splitting (BOS) to improve the weights distribution on the crossbar, thereby enhancing performance potential; 2) inverse-input reusing (IIR) to enhance the sparsity of inputs during inference; and 3) fused scheduling (FS) to reorganize the computation order and leveraging the parallel processing capabilities of the RRAM-based accelerator to improve overall inference efficiency while also reducing data dependencies and storage overhead. Experimental results demonstrate that our proposed methods achieve a $3.44\times $ – $5.53\times $ performance improvement and 55.6%–67.1% energy savings compared with the state-of-the-art RRAM-based accelerator, while also reducing storage overhead by 66.7%–99.1%.

  • Open Access

    A Lifetime Extension Framework for Communication-Intensive Systems Based on Wavelength-Routed Optical Networks-on-Chip

    by Zhidan Zheng, Liaoyuan Cheng, Jeng-De Chang, Tsun-Ming Tseng, Ing-Chao Lin, Ulf Schlichtmann


  • Efficient Model Switching in RRAM-Based DNN Accelerators

    by Fang-Yi Gu, Ing-Chao Lin, Bing Li, Ulf Schlichtmann, Grace Li Zhang

    Resistive random access memory (RRAM) has emerged as a promising technology for deep neural network (DNN) accelerators, but programming every weight in a DNN onto RRAM cells for inference can be both time-consuming and energy-intensive, especially when switching between different DNN models. This article introduces a hardware-aware multimodel merging (HA3M) framework designed to minimize the need for reprogramming by maximizing weight reuse, while taking into account the hardware constraints of the accelerator. The framework includes three key approaches: 1) crossbar (XB)-aware model mapping (XAMM); 2) block-based layer matching (BLM); and 3) multimodel retraining (MMR). XAMM reduces the XB usage of the preprogrammed model on RRAM XBs while preserving the model’s structure. BLM reuses preprogrammed weights in a block-based manner, ensuring the inference process remains unchanged. MMR then equalizes the block-based matched weights across multiple models. Experimental results show that the proposed framework significantly reduces programming cycles in multi-DNN switching scenarios while maintaining or even enhancing accuracy, and eliminating the need for reprogramming.


  • AttentionRC: A Novel Approach to Improve Locality Sensitive Hashing Attention on Dual-Addressing Memory

    by Chun-Lin Chu, Yun-Chih Chen, Wei Cheng, Ing-Chao Lin, Yuan-Hao Chang

    Attention is a crucial component of the Transformer architecture and a key factor in its success. However, it suffers from quadratic growth in time and space complexity as input sequence length increases. One popular approach to address this issue is the Reformer model, which uses locality-sensitive hashing (LSH) attention to reduce computational complexity. LSH attention hashes similar tokens in the input sequence to the same bucket and attends tokens only within the same bucket. Meanwhile, a new emerging nonvolatile memory (NVM) architecture, row column NVM (RC-NVM), has been proposed to support row- and column-oriented addressing (i.e., dual addressing). In this work, we present AttentionRC, which takes advantage of RC-NVM to further improve the efficiency of LSH attention. We first propose an LSH-friendly data mapping strategy that improves memory write and read cycles by 60.9% and 4.9%, respectively. Then, we propose a sort-free RC-aware bucket access and a swap strategy that utilizes dual-addressing to reduce 38% of the data access cycles in attention. Finally, by taking advantage of dual-addressing, we propose transpose-free attention to eliminate the transpose operations that were previously required by the attention, resulting in a 51% reduction in the matrix multiplication time.


  • Open Access

    BasisN: Reprogramming-Free RRAM-Based In-Memory-Computing by Basis Combination for Deep Neural Networks

    by Amro Eldebiky, Grace Li Zhang, Xunzhao Yin, Cheng Zhuo, Ing-Chao Lin, Ulf Schlichtmann, Bing Li


  • A Hardware Friendly Variation-Tolerant Framework for RRAM-Based Neuromorphic Computing

    by Fang-Yi Gu, Cheng-Han Yang, Ing-Chao Lin, Da-Wei Chang, Darsen D. Lu, Ulf Schlichtmann

    Emerging resistive random access memory (RRAM) attracts considerable interest in computing-in-memory by its high efficiency in multiply-accumulate operation, which is the key computation in the neural network (NN). However, due to the imperfect fabrication, RRAM cells suffer from the variations, which make the values in RRAM cells deviate from the target values so that the accuracy of the RRAM-based NN accelerator degrades significantly. Moreover, in a practical hardware design of RRAM-based NN accelerators, if the number of wordlines and bitlines in a crossbar array activated at the same time increases, ADCs with a high resolution are required and the power consumption of ADC increases. This paper proposes a novel methodology to mitigate the impact of variations in RRAM-based neural network accelerators. The methodology includes a unary-based non-uniform quantization method and a variation-aware operation unit (OU) based framework. The unary-based non-uniform quantization method equalizes the significance of weights stored in each RRAM cell to reduce the impact of variations. The variation-aware OU-based framework activates only RRAM cells in the same OU at the same time, which reduces the power consumption of ADCs. Additionally, the framework introduces three methods, including OU skipping, OU recombination, and OU compensation, to further mitigate the impact of variations. The experiments show that the proposed approach outperforms the state-of-the-art among four NN models on two datasets with 2-bit cell resolution.


  • Open Access

    Class-Aware Pruning for Efficient Neural Networks

    by Mengnan Jiang, Jingcun Wang, Amro Eldebiky, Xunzhao Yin, Cheng Zhuo, Ing-Chao Lin, Grace Li Zhang

    Deep neural networks (DNNs) have demonstrated remarkable success in various fields. However, the large number of floating-point operations (FLOPs) in DNNs poses challenges for their deployment in resource-constrained applications, e.g., edge devices. To address the problem, pruning has been introduced to reduce the computational cost in executing DNNs. Previous pruning strategies are based on weight values, gradient values and activation outputs. Different from previous pruning solutions, in this paper, we propose a class-aware pruning technique to compress DNNs, which provides a novel perspective to reduce the computational cost of DNNs. In each iteration, the neural network training is modified to facilitate the class-aware pruning. Afterwards, the importance of filters with respect to the number of classes is evaluated. The filters that are only important for a few number of classes are removed. The neural network is then retrained to compensate for the incurred accuracy loss. The pruning iterations end until no filter can be removed anymore, indicating that the remaining filters are very important for many classes. This pruning technique outperforms previous pruning solutions in terms of accuracy, pruning ratio and the reduction of FLOPs. Experimental results confirm that this class-aware pruning technique can significantly reduce the number of weights and FLOPs, while maintaining a high inference accuracy.

  • Aging-Aware Energy-Efficient Task Deployment of Heterogeneous Multicore Systems

    by Yu-Guang Chen, Chieh-Shih Wang, Ing-Chao Lin, Zheng-Wei Chen, Ulf Schlichtmann

    Heterogeneous multicore systems, which consist of high-performance and power-efficient cores, are emerging to satisfy the various demands on performance and power consumption. On the other hand, as CMOS technology continues to shrink in size, the aging effect, which can cause performance degradation or timing failures, has become a non-negligible threat to lifetime reliability. To overcome the challenges under the aging effect, various approaches have been proposed in previous studies. Most previous studies, however, did not consider the different characteristics of big and little cores. In addition, most of them do not consider critical tasks with the strict timing requirements present in real-time applications, resulting in early system failure. Therefore, considering different characteristics of cores and the presence of critical tasks, we propose an aging-aware task deployment framework for real-time systems. In this framework, for high-performance big cores, we propose a novel asymmetric aging-aware strategy. This strategy finds an energy-efficient task-to-core assignment to reserve some healthy cores at the early system life stage. The reserved cores are kept idle with the lowest voltage and can execute critical tasks at the late system life stage, extending the system lifetime. Meanwhile, the nonreserved cores use lower voltages to execute tasks, reducing the aging effect. For energy-efficient little cores, we adopt the symmetric aging-aware strategy to balance out the aging effect of each little core. With a balanced aging effect, the utilization of little cores is improved. In addition, we propose voltage/frequency boosting and task migration techniques to increase the number of cores that can meet the task timing constraints. Compared to the state of the art, the proposed framework can achieve $1.10\times $ lifetime improvement and 5% energy reduction.


  • An Efficient Implementation of Convolutional Neural Network With CLIP-Q Quantization on FPGA

    by Wei Cheng, Ing-Chao Lin, Yun-Yang Shih

    Convolutional neural networks (CNNs) have achieved tremendous success in the computer vision domain recently. The pursue for better model accuracy drives the model size and the storage requirements of CNNs as well as the computational complexity. Therefore, Compression Learning by InParallel Pruning-Quantization (CLIP-Q) was proposed to reduce a vast amount of weight storage requirements by using a few quantized segments to represent all weights in a CNN layer. Among various quantization strategies, CLIP-Q is suitable for hardware accelerators because it reduces model size significantly while maintaining the full-precision model accuracy. However, the current CLIP-Q approach did not consider the hardware characteristics and it is not straightforward when mapped to a CNN hardware accelerator. In this work, we propose a software-hardware codesign platform that includes a modified version of CLIP-Q algorithm and a hardware accelerator, which consists of $5\times 5$ reconfigurable convolutional arrays with input and output channel parallelization. Additionally, the proposed CNN accelerator maintains the same accuracy of a full-precision CNN in Cifar-10 and Cifar-100 datasets.