Inspired by the great success of neural networks, graph convolutional neural networks (GCNs) are proposed to analyze graph data. GCNs mainly include two phases with distinct execution patterns. The Aggregation phase, behaves as graph processing, showing a dynamic and irregular execution pattern. The Combination phase, acts more like the neural networks, presenting a static and regular execution pattern. The hybrid execution patterns of GCNs require a design that alleviates irregularity and exploits regularity. Moreover, to achieve higher performance and energy efficiency, the design needs to leverage the high intra-vertex parallelism in Aggregation phase, the highly reusable inter-vertex data in Combination phase, and the opportunity to fuse phase-by-phase execution introduced by the new features of GCNs. However, existing architectures fail to address these demands. In this work, we first characterize the hybrid execution patterns of GCNs on Intel Xeon CPU. Guided by the characterization, we design a GCN accelerator, HyGCN, using a hybrid architecture to efficiently perform GCNs. Specifically, first, we build a new programming model to exploit the fine-grained parallelism for our hardware design. Second, we propose a hardware design with two efficient processing engines to alleviate the irregularity of Aggregation phase and leverage the regularity of Combination phase. Besides, these engines can exploit various parallelism and reuse highly reusable data efficiently. Third, we optimize the overall system via inter-engine pipeline for inter-phase fusion and priority-based off-chip memory access coordination to improve off-chip bandwidth utilization. Compared to the state-of-the-art software framework running on Intel Xeon CPU and NVIDIA V100 GPU, our work achieves on average 1509× speedup with 2500× energy reduction and average 6.5× speedup with 10× energy reduction, respectively.

Graph analytics is an emerging application which extracts insights by processing large volumes of highly connected data, namely graphs. The parallel processing of graphs has been exploited at the algorithm level, which in turn incurs three irregularities onto computing and memory patterns that significantly hinder an efficient architecture design. Certain irregularities can be partially tackled by the prior domain-specific accelerator designs with well-designed scheduling of data access, while others remain unsolved.

This work uses patents as agent and proposes the concept of knowledge adaptation factor in view of the objective law of survival of the fittest in technological evolutionary process. Knowledge genetic decomposition of a patent citation network is executed based on the Mendel law. And the contributions of early patents on current patents along patent citations are quantified to evaluate knowledge adaptability. Then the feasibility and effectiveness of the knowledge adaptability evaluation method are verified with the case study of fuel cell. This study collects 3054 patents related to fuel cell which were granted between 1873 to 2002 from the database of United States Patent and Trademark Office, queries from the patent data set of National Bureau of Economic Research to get 14214 patent citations, and uses the network analysis software Pajek to build the patent citation network. The results of the knowledge adaptability evaluation of fuel cell reflect that the knowledge dissemination process is composed of knowledge inheritance, knowledge innovation and knowledge recombinant. The patents whose knowledge adaptation factors are relatively high are subordinate to the technological evolutionary process of fuel cell. It is consistent with the central idea of knowledge adaptation factor, which is technological-evolution-oriented.