XZXin Zhang

Xin Zhang

*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 Zhang has a total of 21 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Preliminaries

    by Shihong Gao, Xin Zhang

  • Open Access

    Apt-Serve: Adaptive Request Scheduling on Hybrid Cache for Scalable LLM Inference Serving

    by Shihong Gao, Xin Zhang, Yanyan Shen, Lei Chen

    Large language model (LLM) inference serving systems are essential to various LLM-based applications. As demand for LLM services continues to grow, scaling these systems to handle high request rates while meeting latency Service-Level Objectives (SLOs), referred to as effective throughput, becomes critical. However, existing systems often struggle to improve effective throughput, primarily due to a significant decline in Time To First Token (TTFT) SLO attainment. We identify two major causes of this bottleneck: (1) memory-intensive KV cache that limits batch size expansion under GPU memory constraints, and (2) rigid batch composition enforced by the default First-Come-First-Serve scheduling policy. In this paper, we introduce Apt-Serve, a scalable framework designed to enhance effective throughput in LLM inference serving. Apt-Serve features a new hybrid cache scheme that combines KV cache with a memory-efficient hidden cache for reusable input hidden state vectors, allowing large batch sizes and improving request concurrency. Based on the hybrid cache, Apt-Serve employs an adaptive runtime scheduling mechanism that dynamically optimizes batch composition. We formally define the adaptive scheduling optimization problem and propose an efficient algorithm with theoretical guarantees. Extensive evaluations on three real-world datasets and LLMs ranging from 13B to 66B parameters demonstrate that Apt-Serve achieves up to 8.8x improvement in effective throughput compared to the state-of-the-art inference serving systems.


  • Robust Graph Neural Networks for Stability Analysis in Dynamic Networks

    by Xin Zhang, Zhen Xu, Yue Liu, Mengfang Sun, Tong Zhou, Wenying Sun

    In the current context of accelerated globalization and digitalization, the complexity and uncertainty of financial markets are increasing, and the identification and prevention of economic risks have become a key link in maintaining the stability of the financial system. Traditional risk identification methods often have limitations because they are difficult to cope with the multi-level and dynamically changing complex relationships in financial networks. With the rapid development of financial technology, graph neural network (GNN) technology, as an emerging deep learning method, has gradually shown great potential in the field of financial risk management. GNN can map transaction behaviors, financial institutions, individuals and their interactive relationships in financial networks into graph structures, and effectively capture potential patterns and abnormal signals in financial data through embedded representation learning. Using this technology, financial institutions can extract valuable information from complex transaction networks, identify hidden dangers or abnormal behaviors that may cause systemic risks in a timely manner, optimize decision-making processes, and improve the accuracy of risk warnings. This paper explores the economic risk identification algorithm based on the GNN algorithm, aiming to provide financial institutions and regulators with more intelligent technical tools to help maintain the security and stability of the financial market. Improving the efficiency of economic risk identification through innovative technical means is expected to further enhance the risk resistance of the financial system and lay the foundation for building a robust global financial system.


  • Open Access

    SIMPLE: Efficient Temporal Graph Neural Network Training at Scale with Dynamic Data Placement

    by Shihong Gao, Yiming Li, Xin Zhang, Yanyan Shen, Yingxia Shao, Lei Chen

    Dynamic graphs are essential in real-world scenarios like social media and e-commerce for tasks such as predicting links and classifying nodes. Temporal Graph Neural Networks (T-GNNs) stand out as a prime solution for managing dynamic graphs, employing temporal message passing to compute node embeddings at specific timestamps. Nonetheless, the high CPU-GPU data loading overhead has become the bottleneck for efficient training of T-GNNs over large-scale dynamic graphs. In this work, we present SIMPLE, a versatile system designed to address the major efficiency bottleneck in training existing T-GNNs on a large scale. It incorporates a dynamic data placement mechanism, which maintains a small buffer space in available GPU memory and dynamically manages its content during T-GNN training. SIMPLE is also empowered by systematic optimizations towards data processing flow. We compare SIMPLE to the state-of-the-art generic T-GNN training system TGL on four large-scale dynamic graphs with different underlying T-GNN models. Extensive experimental results show that SIMPLE effectively cuts down 80.5% ~ 96.8% data loading cost, and accelerates T-GNN training by 1.8× ~ 3.8× (2.6× on average) compared to TGL.


  • Open Access

    Collaborative Graph Neural Networks for Attributed Network Embedding

    by Qiaoyu Tan, Xin Zhang, Xiao Huang, Hao Chen, Jundong Li, Xia Hu

    Graph neural networks (GNNs) have shown prominent performance on attributed network embedding. However, existing efforts mainly focus on exploiting network structures, while the exploitation of node attributes is rather limited as they only serve as node features at the initial layer. This simple strategy impedes the potential of node attributes in augmenting node connections, leading to limited receptive field for inactive nodes with few or even no neighbors. Furthermore, the training objectives (i.e., reconstructing network structures) of most GNNs also do not include node attributes, although studies have shown that reconstructing node attributes is beneficial. Thus, it is encouraging to deeply involve node attributes in the key components of GNNs, including graph convolution operations and training objectives. However, this is a nontrivial task since an appropriate way of integration is required to maintain the merits of GNNs. To bridge the gap, in this paper, we propose COllaborative graph Neural Networks--CONN, a tailored GNN architecture for attribute network embedding. It improves model capacity by 1) selectively diffusing messages from neighboring nodes and involved attribute categories, and 2) jointly reconstructing node-to-node and node-to-attribute-category interactions via cross-correlation. Experiments on real-world networks demonstrate that CONN excels state-of-the-art embedding algorithms with a great margin.

  • Top Cited Papers
    Open Access

    DUCATI: A Dual-Cache Training System for Graph Neural Networks on Giant Graphs with the GPU

    by Xin Zhang, Yanyan Shen, Yingxia Shao, Lei Chen

    Recently Graph Neural Networks (GNNs) have achieved great success in many applications. The mini-batch training has become the de-facto way to train GNNs on giant graphs. However, the mini-batch generation task is extremely expensive which slows down the whole training process. Researchers have proposed several solutions to accelerate the mini-batch generation, however, they (1) fail to exploit the locality of the adjacency matrix, (2) cannot fully utilize the GPU memory, and (3) suffer from the poor adaptability to diverse workloads. In this work, we propose DUCATI, aDual-Cache system to overcome these drawbacks. In addition to the traditionalNfeat-Cache, DUCATI introduces a newAdj-Cache to further accelerate the mini-batch generation and better utilize GPU memory. DUCATI develops a workload-awareDual-Cache Allocator which adaptively finds the best cache allocation plan under different settings. We compare DUCATI with various GNN training systems on four billion-scale graphs under diverse workload settings. The experimental results show that in terms of training time, DUCATI can achieve up to 3.33 times speedup (2.07 times on average) compared to DGL and up to 1.54 times speedup (1.32 times on average) compared to the state-of-the-artSingle-Cache systems. We also analyze the time-accuracy trade-offs of DUCATI and four state-of-the-art GNN training systems. The analysis results offer users some guidelines on system selection regarding different input sizes and hardware resources.


  • Open Access

    Occurrence of a Likely Tuff Bed between the Middle and Upper Siwaliks, Taunsa area, Dera Ghazi Khan, Eastern Sulaiman Range, Pakistan

    by Rahman Ullah, Nie Fengjuin, Zhang Chengyong, Saqib Izhar, Idrees Safdar, Zhang Xin, Asim Ali

    A likely tuff bed lies along the gradational contact of the Middle and Upper Siwaliks in eastern Sulaiman Range, Taunsa area of Dera Ghazi Khan district, Pakistan. This tuffaceous unit is 0.5–3 m thick and extends for 10 km along the north-south strike in the eastern limb of the Zindapir anticline. It is greyish white to white on fresh surface, fine-grained to silty at the bottom and clayey at the top and thus shows a fining upward grain-size grading. The lower part of the ash bed shows a prominent lamination defined by megascopically visible abundant biotite, while the central and upper parts are so fine-grained that the individual minerals cannot be seen in hand sample. Unlike the lower well-laminated part, the central and upper parts are crudely laminated to apparently massive. The bulk samples analysed with X-ray diffraction consist of quartz, feldspar (plagioclase), biotite, clays, calcite and some ore mineral likely spinel, while the clay-size fractions contain illite, chlorite, biotite and probably their mixed-layered varieties. The colour, texture, presence of abundant biotite and stratigraphic position of the Taunsa tuff correlate with those reported from Potwar plateau and from Kashmir basin. However, the apparent absence of smectite from the XRD pattern makes the Taunsa ash bed different from both Potwar and Kashmir tuffs. The present stratigraphic position of the tuff bed corresponds to shallow diagenetic zone, while the absence of smectite in the tuff and crystallinity of illite suggest that the tuff is probably derived upon reworking from a deeper diagenetic zone belonging to a lower stratigraphic level. The Eocene or other older pre-Siwalik units in Pakistan may have or had some primary ashfall deposits as reported in the northwestern Himalayas of India. This older volcanic ash may have been reworked to its present site of occurrence along the gradational contact of the Middle and the Upper Siwaliks in Taunsa area of Dera Ghazi Khan. However, the primary source of the Taunsa tuff may belong more likely to Chagai arc in Pakistan than to Dacht-e-Nawar volcanic complex in Afghanistan.