Modern AI data centers face a critical paradox: while machine learning workloads dominate infrastructure demands, actual GPU utilization remains consistently low. Existing schedulers fail to coordinate heterogeneous resources effectively, lack predictive capabilities for dynamic workloads, and cannot balance isolation requirements with sharing optimization in multi-tenant clusters. This paper presents Wind, a novel resource scheduler that bridges the GPU utilization gap through predictive scheduling and geometric resource coordination. Wind introduces three key innovations: (1) a resource prediction framework that leverages historical execution patterns to forecast task requirements and completion times with high accuracy;(2) a unified scheduling architecture supporting isolation, sharing, preemption, and prioritization policies that eliminate resource fragmentation while maintaining performance guarantees; and (3) a Hilbert curve-based multi-dimensional scheduling algorithm that maps CPU-memory-GPU resource space to preserve spatial locality while achieving linear computational complexity.

Due to the escalating demand to analyze large graphs, many organizations are now collecting billion-level property graph datasets, concurrently executing many complex graph queries against them, and expecting interactive-level response latency. However, such requirements are particularly challenging because of the notoriously irregular data access pattern and complex dependencies between heterogeneous subtasks. Despite the widespread availability of many-core CPUs and high-speed networking in modern datacenters, existing distributed graph query systems struggle with their inherent inefficiencies, resulting in low hardware utilization and poor query performance on these state-of-the-art hardware. To address these challenges, we introduce the Partitioned Stateful Traversal Machine (PSTM), which extends the Gremlin graph traversal machine. PSTM retains the expressive power of the Gremlin query language, enabling it to accommodate a wide range of graph query tasks, including traversal, pattern matching, filtering, and result aggregation. It additionally introduces query memoranda, allowing for more efficient implementation and execution of numerous graph queries in distributed environments. Moreover, PSTM facilitates various system-level optimizations, such as massively parallel execution, overlapping computation with communication, locality-aware data access, and lightweight progress tracking. Building upon PSTM, we develop GraphDance, a distributed graph database featuring an efficient asynchronous PSTM run-time. Our evaluations, conducted on an 8-node cluster, show that GraphDance achieves millisecond-level query latency for complex queries on terabyte-scale graphs, with an average latency reduction of 89.2% across all interactive complex queries in the LDBC SNB benchmark compared to existing distributed graph query systems.

Modern stream processing engines (SPEs) must handle massive real-time data streams under strict latency and throughput requirements. However, conventional SPEs are constrained by their software parallelization strategies (e.g., queue-based data re-partitioning, high synchronization overheads, etc.), which prevent efficient utilization of modern hardware capabilities, ultimately limiting performance scalability. In this paper, we present FlexStream, a novel SPE that leverages hardware offloading to redesign the parallelization strategies and overcome these limitations. By offloading data re-partitioning to hardware and integrating a coupled network-executor model, FlexStream maximizes resource utilization, achieving up to 95% network bandwidth saturation. To address the load imbalance challenges introduced by this design, we implement a lock-free state backend with efficient state migration mechanisms. Overall, FlexStream achieves throughput improvements of 1.95 × - 3.35 × compared to state-of-the-art SPEs (e.g., LightSaber) across six real-world streaming analytics applications. FlexStream cuts latency spikes by 71.9% and migration time by 66.8% during state migration, highlighting the benefits of hardware-software co-design in SPEs. Our work underscores the potential of hardware-software co-design in SPEs, offering a scalable, elastic solution for real-time analytics.

Variable-Length Graph Pattern Matching (VLGPM) is a critical functionality in graph databases, pivotal for identifying patterns where the number of connecting edges between two matched vertices is variable. This function plays a vital role in analyzing complex and dynamic networks such as social networks or bank transfers networks, where relationships can vary extensively in both length and structure. However, despite its importance, current graph databases, optimized primarily for single-hop subgraph matching, struggle with VLGPM over large graphs.

Serverless computing is renowned for its computation elasticity, yet its full potential is often constrained by the requirement for functions to operate within local and dedicated background environments, resulting in limited memory elasticity. To address this limitation, this paper introduces TrEnv, a co-designed integration of the serverless platform with the operating system and CXL/RDMA-based remote memory pools in two key areas. Firstly, TrEnv introduces repurposable sandboxes, which can be shared across different functions and hence, substantially decrease the overhead associated with creating isolation sandboxes. Secondly, it augments the OS with "memory templates" that enable rapid restoration of function states stored on remote memory. These innovations allow TrEnv to facilitate rapid transitions between instances of different functions and enable memory sharing across multiple nodes. Our evaluations using a variety of representative and real-world workloads demonstrate that TrEnv can initiate a container within 10 milliseconds, achieving up to a 7× speedup in P99 end-to-end latency and reducing memory usage by 48% on average compared to state-of-the-art on-demand restoring systems.

Out-of-core random walk system has recently attracted a lot of attention as an economical way to run billions of walkers over large graphs. However, existing out-of-core random walk systems are all built upon general out-of-core graph processing frameworks, and hence do not take advantage of the unique properties of random walk applications. Different from traditional graph analysis algorithms, the sampling process of random walk can be decoupled from the processing of the walkers. It enables the system to reserve only pre-sample results in memory, which are typically much smaller than the entire edge set. Moreover, in random walk, it is not the number of walkers but the number of steps moved per second that dominates the overall performance. Thus, with independent walkers, there is no need to process all the walkers simultaneously.

Many real-world graphs are temporal in nature, where the temporal information indicates when a particular edge is changed (e.g., edge insertion and deletion). Performing random walks on such temporal graphs is of paramount value. The state-of-the-art sampling strategies are tailored for conventional static graphs and thus cannot effectively tackle the dynamic nature of temporal graphs due to several significant efficiency challenges, i.e., high sampling complexity, gigantic index space, and poor programmability.

The efficiency of distributed shared memory (DSM) has been greatly improved by recent hardware technologies. But, the difficulty of distributed memory management can still be a major obstacle to the democratization of DSM, especially when a partial failure of the participating clients (e.g., due to crashed processes or machines) should be tolerated.

As many real-world applications are streaming and attached with time instances, a few works have been proposed to learn streaming graph neural networks (GNNs). Unfortunately, current streaming GNNs are observed to have a large training overhead and suffer from bad parallel scalability on multiple GPUs. These drawbacks pose severe challenges to online learning of streaming GNNs and their application to real-time scenarios. To improve training efficiency, one promising solution is to use sampling, a technique widely used in static GNNs. However, to the best of our knowledge, sampling has not been investigated in learning streaming GNNs. Based on these observations, in this paper, we propose T-GCN, the first sampling-based streaming GNN system, which targets temporal-aware streaming graphs and takes advantage of a hybrid CPU-GPU co-processing architecture to achieve high throughput and low latency. T-GCN proposes an efficient sampling method, namely Segment Its Search, to offer high sampling speed with respect to three typical types of general graph sampling methods (i.e., node-wise, layer-wise, and subgraph sampling). We propose a locality-aware data partitioning method to reduce CPU-GPU communication latency and data transfer overhead, and an NVLink-specific task schedule to fully exploit NVLink's fast speed and improve GPU-GPU communication efficiency. Besides, we further pipeline the computation and the communication by introducing an efficient memory management mechanism, to improve scalability while hiding data communication. Overall, with respect to end-to-end performance, for single-GPU training, T-GCN achieves up to 7.9× speedup than state-of-the-art works. In terms of scalability, T-GCN runs 5.2× faster on average with 8 GPUs than one GPU. Additionally, in terms of sampling, T-GCN also yields a maximum of 38.8× speedup with our Segment Its Search sampling method.

Temporal graphs attach time information to edges and are commonly used for implementing time-critical applications that can not be effectively processed by traditional static and dynamic graph processing engines. State-of-the-art solutions that target temporal path problems remain ad-hoc and often suboptimal. A unified and high-performance solution that could efficiently process general temporal path problems via a universal optimization strategy and relieve practitioners from heavy optimization efforts is in urgent demand. In this paper, we make two key observations: (1) temporal path problems can be described as topological-optimum problems and solved by a universal single scan execution model; and (2) data redundancy commonly occurs in the native format of the transformed temporal graphs, which is unnecessary for information propagation and can be eliminated for better memory utilization and execution efficiency. Based on these core insights, we propose TegRaph, the first general-purpose temporal graph computing engine to provide a unified optimization strategy and execution model for general temporal path problems and their applications. TegRaph not only presents temporal information-aware graph representation that naturally fits temporal graphs but also offers general system-level supports such as out-of-core execution. Extensive evaluation reveals that TegRaph can achieve significant speedups over the state-of-the-art designs with up to two orders of magnitude (241×) with the throughput of two hundred million edges per second.

Nowadays, it is a rapidly rising demand yet challenging issue to run large-scale applications on shared infrastructures such as data centers and clouds with low execution latency and high resource utilization. This paper reports our experience with Swift, a system capable of efficiently running real-time and interactive data processing jobs at cloud scale. Taking directed acyclic graph DAG as the job model, Swift achieves the design goal by three new mechanisms: 1 fine-grained scheduling that can efficiently partition a job into graphlets i.e., sub-graphs based on new shuffle heuristics and that does scheduling in the unit of graphlet, thus avoiding resource fragmentation and waste, 2 adaptive memory-based in-network shuffling that reduces IO overhead and data transfer time by doing shuffle in memory and allowing jobs to select the most efficient way to fulfill shuffling, and 3 lightweight fault tolerance and recovery that only prolong the whole job execution time slightly with the help of timely failure detection and fine-grained failure recovery. Experimental results show that Swift can achieve an average speedup of 2.11× on TPC-H, and 14.18× on Terasort when compared with Spark. Swift has been deployed in production, supporting as many as 140,000 executors and processing millions of jobs per day. Experiments with production traces show that Swift outperforms JetScope and Bubble Execution by 2.44× and 1.23× respectively.

360-degree video streaming, which is becoming more and more popular as the fast development of VR/AR applications nowadays due to the immersive viewing experience it can offer, poses enormous challenges to the current network infrastructure in terms of high bandwidth and low latency requirements. To address this problem and to ensure the QoE (quality of experience) of end-users, this paper presents CUBIST, a method and system for high-quality 360-degree video streaming in networks with cache nodes at the edge. To the best of our knowledge, it is the first tile-based edge caching solution that incorporates proactive tile prefetching and hierarchical cache organization into reactive caching to maximize the caching benefit while reducing the cost of 360-degree video streaming. Experimental results show that CUBIST can achieve a cache hit ratio of 87 % and improve the effective video bitrate by 12.9 % with most rate transitions being small when compared with the latest FoV-aware edge caching scheme.

Caching contents in a small cell base station (SBS) is getting supported more and more widely today due to the Internet traffic growth and the requirement of low access latency. A primary concern and challenging issue with cache-enabled SBSs is how to better utilize network resources to achieve high overall performance. Though existing caching strategies can solve the problem to some extent, they are far from perfect --- pure popularity-based ones usually lead to sub-optimal caching performance whereas global coordination ones suffer from extra cost of many control messages. To deal with the issue, we present an adaptive caching scheme based on Bayesian inference, which 1) identifies the traffic features over time for each SBS; 2) synthesizes various features to rank the contents via a Bayesian ranking model; and 3) does cache placement online according to the ranking results. Unlike existing approaches that only highlight some specific factor, our scheme, due to the adoption of a Bayesian approach, can easily support additional features of high impact on caching performance and measure them in a decentralized way within a single SBS. We evaluate our scheme under various circumstances in terms of SBSs density, cache size, content popularity and skewness. The results show that our solution using multiple features exhibits improved performance --- it can reduce more than 30% the overall network latency in some cases when compared with solutions that only use a single feature.

There has been increasing attention to virtual reality applications in recent years, especially to immersive or 360-degree videos that typically consume much more bandwidth than traditional ones. Though all produced data is transferred, only a small part (denoted as Field of View or viewport) is watched by users due to the nature of immersive videos. Obviously, this causes a large waste of network resources. Hence, it is important to define a viewport-dependent streaming transmission strategy by detecting where the user is gazing and the movement of the user's head. Unfortunately, there are few datasets providing this information. In this paper, we propose a tile-based simulation approach to generate the distribution of the user's behavior and to provide information that can be used to optimize future view-dependent streaming protocols. We first characterize the users' viewport pattern from datasets gathered from real users by decomposing the 360-degree stream into tiles and analyzing the frequency and time-interval distribution for each tile. Then, we devise a hierarchical Markov model that incorporates the beta distribution of each tile time interval to predict tile transition. The results show that the simulation tool characterizes the tile sequences of users accurately, performing close to the empirical results.

This paper reports our efforts on swCaffe, a highly efficient parallel framework for accelerating deep neural networks (DNNs) training on Sunway TaihuLight, the current fastest supercomputer in the world that adopts a unique many-core heterogeneous architecture, with 40,960 SW26010 processors connected through a customized communication network.First, we point out some insightful principles to fully exploit the performance of the innovative many-core architecture.Second, we propose a set of optimization strategies for redesigning a variety of neural network layers based on Caffe.Third, we put forward a topology-aware parameter synchronization scheme to scale the synchronous Stochastic Gradient Descent (SGD) method to multiple processors efficiently.We evaluate our framework by training a variety of widely used neural networks with the ImageNet dataset.On a single node, swCaffe can achieve 23%˜119% overall performance compared with Caffe running on K40m GPU.As compared with the Caffe on CPU, swCaffe runs 3.04˜7.84xfaster on all the networks.Finally, we present the scalability of swCaffe for training of ResNet-50 and AlexNet on the scale of 1024 nodes.

In-network caching is considered to be an important solution to efficiently using network resources to achieve a high overall content delivery performance in both information-centric networks (ICNs) and 5G wireless networks. Content placement plays a key role in achieving this goal. Unfortunately, most content placement strategies today rely on opportunistic caching due to the problem complexity. We analyze the content placement problem in detail and present MCPC, a new content placement strategy for architectures that support in-network caching. Unlike existing content placement approaches that try to increase cache hit ratio, MCPC leverages the information recorded in each network node to reduce content access latency. MCPC proposes two new mechanisms: 1) a content load allocation estimation method based on local requests aggregation information; and 2) a content placement algorithm that avoids long-distance signaling messages by partitioning the network into smaller domains. We evaluate MCPC on a variety of network topologies, cache sizes and content popularity distributions. The experimental results show that MPCP can reduce by up to 56% the content access latency while providing comparable cache hit ratios as traditional benchmarks.

The exponential growth of data traffic raises a great challenge to content delivery in current TCP/IP networks. To answer this challenge, Information-Centric Networking (ICN) has been proposed with the purpose of bringing content caching and name-based content access to the network layer. Though great progress has been made, most existing ICN proposals lack support for parallel data transfer over multiple paths with low data redundancy. To deal with the issue, we present CODA, a fully distributed cooperative multipath data transmission solution that enhances content delivery further. Taking Named Data Networking (NDN) as a basis, CODA works in a distributed manner with the following contributions: 1) it extends the standard Interest model in NDN to support transmission of data over multiple paths so as to reduce the flow completion time; 2) it devises a traffic scheduling model to form parallel paths for transmitting data in a cooperative way; and 3) it proposes a transmission control scheme to select paths in an efficient and reliable manner. Extensive simulation comparisons with existing data transmission methods show that: 1) CODA speeds up the data rate twice as high as that of the best-route method; and 2) the amount of Interests required by CODA to build multiple data transmission paths in the network accounts for only 66% of that by MSRT, another multipath transmission proposal.

MapReduce, as a programming model and implementation for processing large data sets on clusters with hundreds or thousands of nodes, has gained wide adoption. In spite of the fact, we found that MapReduce on commodity clusters, which are usually equipped with limited memory and hard-disk drive (HDD) and have processors of multiple or many cores, does not scale as expected as the number of processor cores increases. The key reason for this is that the underlying low-speed HDD storage cannot meet the requirement of frequent IO operations. Though in-memory caching can improve IO, it is costly and sometimes cannot get the desired result either due to memory limitation. To deal with the problem and make MapReduce more scalable on commodity clusters, we present mpCache, a solution that utilizes solid-state drive (SSD) to cache input data and localized data of MapReduce tasks. In order to make a good trade-off between cost and performance, mpCache proposes ways to dynamically allocate the cache space between the input data and localized data and to do cache replacement. We have implemented mpCache in Hadoop and evaluated it on a 7-node commodity cluster by 13 benchmarks. The experimental results show that mpCache can gain an average speedup of 2.09× when compared with Hadoop, and can achieve an average speedup of 1.79× when compared with PACMan, the latest in-memory optimization of MapReduce.

As a widely used programming model and implementation for processing large data sets, MapReduce performs poorly on heterogeneous clusters, which, unfortunately, are common in current computing environments. To deal with the problem, this paper: 1) analyzes the causes of performance degradation and identifies the key one as the large volume of inter-node data transfer resulted from even data distribution among nodes of different computing capabilities, and 2) proposes ActCap, a solution that uses a Markov chain based model to do node-capability-aware data placement for the continuously incoming data. ActCap has been incorporated into Hadoop and evaluated on a 24-node heterogeneous cluster by 13 benchmarks. The experimental results show that ActCap can reduce the percentage of inter-node data transfer from 32.9% to 7.7% and gain an average speedup of 49.8% when compared with Hadoop, and achieve an average speedup of 9.8% when compared with Tarazu, the latest related work.

A virtual machine (VM) has been serving as a crucial component in cloud computing with its rich set of convenient features. The high overhead of a VM has been well addressed by hardware support such as Intel virtualization technology (VT), and by improvement in recent hypervisor implementation such as Xen, KVM, etc. However, the high demand on VM image storage remains a challenging problem. Existing systems have made efforts to reduce VM image storage consumption by means of deduplication within a storage area network (SAN) cluster. Nevertheless, an SAN cannot satisfy the increasing demand of large-scale VM hosting for cloud computing because of its cost limitation. In this paper, we propose Liquid, a scalable deduplication file system that has been particularly designed for large-scale VM deployment. Its design provides fast VM deployment with peer-to-peer (P2P) data transfer and low storage consumption by means of deduplication on VM images. It also provides a comprehensive set of storage features including instant cloning for VM images, on-demand fetching through a network, and caching with local disks by copy-on-read techniques. Experiments show that Liquid's features perform well and introduce minor performance overhead.

As the evolution of applications, software systems become more and more complex. However, it is a challenge to deliver easy-to-use and reliable large-scale distributed software systems. This paper first reviews current requirements on software systems and the progress made and then reports our effort in delivering a more usable and reliable large-scale distributed software development framework (iFrame) to support e-research. Bearing human collaboration in the first place, iFrame synthesizes peer-to-peer (P2P), service-oriented architecture (SOA), Web 2.0 and semantic Web into grid: P2P is deployed to construct an interconnected environment to enhance reliability and scalability, SOA and semantic Web are exploited to deliver concrete business functionality and prompt interoperability, whereas Web 2.0 is utilized to devise more usable interface.

Today's difficult business reality makes enterprises more and more resort to information technology to manage nearly everything within the enterprises and to conduct business activities. This puts forward high requirements on information systems and in turn, raises a great challenge to software vendors - how to develop systems according to the diverse requirements as quickly as possible while spending a cost as low as possible. To achieve this purpose, we argue a core infrastructure delivering some general functions is needed by software vendors. The paper proposes such an infrastructure and details its design philosophy. A scenario is also presented to illustrate the effectiveness and efficiency of the infrastructure proposed.

Context awareness, context sharing and context processing are key requirements for the future CSCW, HCI and ubiquitous computing systems. Until now, the collaborative context factors have been seldom specifically addressed. This paper argues that a generic context model is very important for building context-aware collaborative applications. A new semantic rich context model for collaborative environment is proposed. The conceptual model for context is described as ontology for contextual collaborative applications (OCCA). The model for context query & memory service, context matching service and control policies is illustrated. Information space, interaction space and collaboration control mechanisms are built up or implemented based on this context model

There is an increasing need for computer-supported cooperative work (CSCW) in recent years. However, most of existing collaborative systems are not scalable enough, thus leading to bad usability. In this paper we propose a scalable framework for large-scale distributed collaboration which aims to support a wide range of collaboration requirements. The framework comprises a set of geographically dispersed collaborative servers (co-servers) specially deployed by participant organizations. The set of co-servers constitutes an overlay network which provides an infrastructural supporting environment for large-scale distributed collaboration. The framework is scalable in terms of the geographic distribution of participant organizations, size of groups, and total number of groups. The framework will be used to build the next-generation collaborative e-learning platform. A prototype system is being developed to demonstrate the application of the framework

Transaction concept plays a key role in business success. However, existing transaction models are not applicable in service grid environment. Though many new models have been suggested, they still have some deficiencies, e.g., major extension to services are required, and they are hard to implement, to name a few. In this paper, a new model is proposed based on the analysis of existing work and the characteristics of service grid environment. Besides, model implementation issues are also covered

The harmonic convergence of service-oriented computing and grid results in the emergence of service grid and makes grid technology evolve into business domain. Along this evolution, the problem of transaction support becomes crucial because it is transaction that determines the adoption and success of business grid applications at most time. However, existing transaction models are not applicable in service grid environment. Though many new models have been suggested, they still have some deficiencies. In this paper, a new model is proposed based on the analysis of existing work. Take the behavior of services, the interdependence between them and the high-level requirement on transaction properties as the starting points, the proposed model is easy to implement and suitable for grid environment

The emergence of e-commerce (EC) brings great challenges to both research community and industries. As an enabling technology, workflow is viewed as a good basis for EC solutions due to its immense power in dealing with complex tasks and integration of heterogeneous resources. However, there are many issues with existing workflow systems such as rigid process definition, lack of interoperability, limited support of transaction, to name a few. This paper details how to support EC applications by extending Cova, a meta-groupware with process management ability, to support Web services. Based on the ability of run-time process adjustment provided by Cova, the dynamics of EC applications can be well addressed, while the introduction of Web services facilitates the interaction between different organizations without violating their autonomy.

Though 20 years have passed since the birth of CSCW, the original goal of it is not reached as well as people expected. This situation is mostly due to the supporting technology especially the infrastructure. Today, great changes have taken place in technology, including grid computing and Web services. These technologies, we think, significantly affect the application of CSCW. In this paper, a framework called CoFrame is proposed to answer the challenges faced by CSCW. Based on the emerging grid and Web service technologies, CoFrame provides some general yet flexible cooperation related services and organizes them into different layers. The elaborately designed services and architecture make CoFrame adaptive to diverse requirements of different domains. The paper details the framework and demonstrates its application with a case study in e-learning.

Although they are being exploited by businesses in a variety of industries, current workflow systems are considered to be too rigid. However, it is difficult to construct a new workflow system from scratch, due to its complexity. On the other hand, meta-groupware is attracting more and more attention, in that it can be used to compose sophisticated systems. In this paper, we show how to implement a workflow management system (WfMS) by using Cova, a meta-groupware system developed at Tsinghua University. Problems, including process design, run-time synchronous support and reliability, are addressed. It is efficient as well as easy to construct a WfMS based on meta-groupware. In addition, a system developed in this way presents many desired features, such as flexibility, reliability, and so on.