Partitioning and distributing deep neural networks (DNNs) across end-devices, edge resources and the cloud has a potential twofold advantage: preserving privacy of the input data, and reducing the ingress bandwidth demand beyond the edge. However, for a given DNN, identifying the optimal partition configuration for distributing the DNN that maximizes performance is a significant challenge. This is because the combination of potential target hardware resources that maximizes performance and the sequence of layers of the DNN that should be distributed across the target resources needs to be determined, while accounting for user-defined objectives/constraints for partitioning. This paper presents Scission, a tool for automated benchmarking of DNNs on a given set of target device, edge and cloud resources for determining optimal partitions that maximize DNN performance. The decision-making approach is context-aware by capitalizing on hardware capabilities of the target resources, their locality, the characteristics of DNN layers, and the network condition. Experimental studies are carried out on 18 DNNs. The decisions made by Scission cannot be manually made by a human given the complexity and the number of dimensions affecting the search space. The benchmarking overheads of Scission allow for responding to operational changes periodically rather than in real-time. Scission is available for public download at https://github.com/qub-blesson/Scission.
Industrial Internet of Things (IIoT) applications can benefit from leveraging edge computing. For example, applications underpinned by deep neural networks (DNN) models can be sliced and distributed across the IIoT device and the edge of the network for improving the overall performance of inference and for enhancing privacy of the input data, such as industrial product images. However, low network performance between IIoT devices and the edge is often a bottleneck. In this study, we develop ScissionLite, a holistic framework for accelerating distributed DNN inference using the Transfer Layer …
Northeastern University（东北大学）；MIT Lincoln Laboratory（麻省理工学院林肯实验室）
Deep learning model inference is a key service in many businesses and scientific discovery processes. This paper introduces RIBBON, a novel deep learning inference serving system that meets two competing objectives: quality-of-service (QoS) target and cost-effectiveness. The key idea behind RIBBON is to intelligently employ a diverse set of cloud computing instances (heterogeneous instances) to meet the QoS target and maximize cost savings. RIBBON devises a Bayesian Optimization-driven strategy that helps users build the optimal set of heterogeneous instances for their model inference service …
This paper investigates the existing resource management approaches in Cloud Data Centres for energy and thermal efficiency. It identifies the need for integrated computing and cooling systems management and learning-based solutions in resource management systems. A taxonomy on energy and thermal efficient resource management in data centres is proposed based on an in-depth analysis of the literature. Furthermore, a detailed survey on existing approaches is conducted according to the taxonomy and recent advancements including machine learning-based resource management approaches and cooling ma…
Edge computing is an emerging technology which places computing at the edge of the network to provide an ultra-low latency. Computation offloading, a paradigm that migrates computing from mobile devices to remote servers, can now use the power of edge computing by offloading computation to cloudlets in edge-clouds. However, the task scheduling of computation offloading in edge-clouds faces a two-fold challenge. First, as cloudlets are geographically distributed, it is difficult for each cloudlet to perform load balancing without centralized control. Second, as tasks of computation offloading h…
This paper presents our preliminary results with ABEONA, an edge-to-cloud architecture that allows migrating tasks from low-energy, resource-constrained devices on the edge up to the cloud. Our preliminary results on artificial and real world datasets show that it is possible to execute workloads in a more efficient manner energy-wise by scaling horizontally at the edge, without negatively affecting the execution runtime.
Day after day, the number of mobile applications deployed on cloud computing continues in increasing because o f smartphone capabilities improvement. Cloud computing has already succeeded in the web based application, for that reason, the demand for context aware services provided by cloud computing increases. To customize a cloud service that takes into account th e consumer requirements, which depend on information change, it brings to light many recent challenges to cloud computing about environment aware, location aware, time aware. The cloud provider, moreover, has to manage personalized …
As a Distributed Hash Table (DHT), Skip Graph routing overlays are exploited in several peer-to-peer (P2P) services, including P2P cloud storage. The fully decentralized replication algorithms that are applicable to the Skip Graph-based P2P cloud storage fail on improving the performance of the system with respect to both the availability of replicas as well as their response time. Additionally, they presume the system as homogeneous with respect to the nodes' latency distribution, availability behavior, bandwidth, or storage. In this paper, we propose Pyramid, which is the first fully decentr…
Cloud applications are increasingly distributing data across multiple regions and cloud providers. Unfortunately, wide-area bulk data transfers are often slow, bottlenecking applications. We demonstrate that it is possible to significantly improve inter-region cloud bulk transfer throughput by adapting network overlays to the cloud setting -- that is, by routing data through indirect paths at the application layer. However, directly applying network overlays in this setting can result in unacceptable increases in cloud egress prices. We present Skyplane, a system for bulk data transfer between…
Resource allocation (RA) is a significant aspect in Cloud Computing which facilitates the Cloud resources to Cloud consumers as a metered service. The Cloud resource manager is responsible to assign available resources to the tasks for execution in an effective way that improves system performance, reduce response time, reduce makespan and utilize resources efficiently. To fulfil these objectives, an effective Tasks Scheduling algorithm is required. The standard Min-Min and Max-Min Task Scheduling Algorithms are available, but these algorithms are not able to produce better makespan and effect…
Current cloud computing frameworks host millions of physical servers that utilize cloud computing resources in the form of different virtual machines (VM). Cloud Data Center (CDC) infrastructures require significant amounts of energy to deliver large scale computational services. Computing nodes generate large volumes of heat, requiring cooling units in turn to eliminate the effect of this heat. Thus, the overall energy consumption of the CDC increases tremendously for servers as well as for cooling units. However, current workload allocation policies do not take into account the effect on tem…
In cloud ML inference systems, batching is an essential technique to increase throughput which helps optimize total-cost-of-ownership. Prior graph batching combines the individual DNN graphs into a single one, allowing multiple inputs to be concurrently executed in parallel. We observe that the coarse-grained graph batching becomes suboptimal in effectively handling the dynamic inference request traffic, leaving significant performance left on the table. This paper proposes LazyBatching, an SLA-aware batching system that considers both scheduling and batching in the granularity of individual g…
Distributed data processing platforms (e.g., Hadoop, Spark, and Flink) are widely used to distribute the storage and processing of data among computing nodes of a cloud. The centralization of cloud resources has given birth to edge computing, which enables the processing of data closer to the data source instead of sending it to the cloud. However, due to resource constraints such as energy limitations, edge computing cannot be used for deploying all kinds of applications. Therefore, tasks are offloaded from an edge device to the more resourceful cloud. Previous research has evaluated the ener…
MSAO: Adaptive Modality Sparsity-Aware Offloading with Edge-Cloud Collaboration for Efficient Multimodal LLM Inference
Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算技术研究所）；University of Chinese Academy of Sciences（中国科学院大学）；University of Science and Technology of China（中国科学技术大学）；University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）
Because of the growing interest for mobile device and pervasive applications deployed on cloud computing, the providing of intelligent and ubiquitous context-aware applications that take into account the user's context is one of the main challenges in future applications. In this article we consider how to augment applications aware context used by mobiles device and deployed on cloud computing. The behavior of these applications should depend not only on their internal state and user interactions but also on the context sensed during their execution. Indeed, our approach based on two essentia…
Saving energy is an important issue for cloud providers to reduce energy cost in a data center. With the increasing popularity of cloud computing, it is time to examine various energy reduction methods for which energy consumption could be reduced and lead us to green cloud computing. In this paper, our aim is to propose a virtual machine selection algorithm to improve the energy efficiency of a cloud data center. We are also presenting experimental results of the proposed algorithm in a cloud computing based simulation environment. The proposed algorithm dynamically took the virtual machines'…
Autoscaling system can reconfigure cloud-based services and applications, through various configurations of cloud software and provisions of hardware resources, to adapt to the changing environment at runtime. Such a behavior offers the foundation for achieving elasticity in modern cloud computing paradigm. Given the dynamic and uncertain nature of the shared cloud infrastructure, cloud autoscaling system has been engineered as one of the most complex, sophisticated and intelligent artifacts created by human, aiming to achieve self-aware, self-adaptive and dependable runtime scaling. Yet, exis…
We present a model for measuring the impact of offloading soft real-time jobs over multi-tier cloud infrastructures. The jobs originate in mobile devices and offloading strategies may choose to execute them locally, in neighbouring devices, in cloudlets or in infrastructure cloud servers. Within this specification, we put forward several such offloading strategies characterised by their differential use of the cloud tiers with the goal of optimizing execution time and/or energy consumption. We implement an instance of the model using Jay, a software framework for adaptive computation offloadin…
Database platform-as-a-service (dbPaaS) is developing rapidly and a large number of databases have been migrated to run on the Clouds for the low cost and flexibility. Emerging Clouds rely on the tenants to provide the resource specification for their database workloads. However, they tend to over-estimate the resource requirement of their databases, resulting in the unnecessarily high cost and low Cloud utilization. A methodology that automatically suggests the "just-enough" resource specification that fulfills the performance requirement of every database workload is profitable. To this end,…
University of Washington（华盛顿大学）；Microsoft Research（微软研究院）；Facebook AI（脸书人工智能研究院）
ML workloads are becoming increasingly popular in the cloud. Good cloud training performance is contingent on efficient parameter exchange among VMs. We find that Collectives, the widely used distributed communication algorithms, cannot perform optimally out of the box due to the hierarchical topology of datacenter networks and multi-tenancy nature of the cloudenvironment.In this paper, we present Cloud Collectives , a prototype that accelerates collectives by reordering theranks of participating VMs such that the communication pattern dictated by the selected collectives operation best exploi…
Cloud computing has attracted both end-users and Cloud Service Providers (CSPs) in recent years. Improving resource utilization rate (RUtR), such as CPU and memory usages on servers, while maintaining Quality-of-Service (QoS) is one key challenge faced by CSPs with warehouse-scale data centers. Prior works proposed various algorithms to reduce energy cost or to improve RUtR, which either lack the fine-grained task scheduling capabilities, or fail to take a comprehensive system model into consideration. This article presents H2O-Cloud, a Hierarchical and Hybrid Online task scheduling framework …
Smart farming is a recent innovation in the agriculture sector that can improve the agricultural yield by using smarter, automated, and data driven farm processes that interact with IoT devices deployed on farms. A cloud-fog infrastructure provides an effective platform to execute IoT applications. While fog computing satisfies the real-time processing need of delay-sensitive IoT services by bringing virtualized services closer to the IoT devices, cloud computing allows execution of applications with higher computational requirements. The deployment of IoT applications is a critical challenge …
Today's clusters often have to divide resources among a diverse set of jobs. These jobs are heterogeneous both in execution time and in their rate of arrival. Execution time heterogeneity has lead to the development of hybrid schedulers that can schedule both short and long jobs to ensure good task placement. However, arrival rate heterogeneity, or burstiness, remains a problem in existing schedulers. These hybrid schedulers manage resources on statically provisioned cluster, which can quickly be overwhelmed by bursts in the number of arriving jobs. In this paper we propose CloudCoaster, a hyb…
Recent years have witnessed a rapid growth of deep-network based services and applications. A practical and critical problem thus has emerged: how to effectively deploy the deep neural network models such that they can be executed efficiently. Conventional cloud-based approaches usually run the deep models in data center servers, causing large latency because a significant amount of data has to be transferred from the edge of network to the data center. In this paper, we propose JALAD, a joint accuracy- and latency-aware execution framework, which decouples a deep neural network so that a part…
Software-defined internet of vehicles (SDIoV) has emerged as a promising paradigm to realize flexible and comprehensive resource management, for next generation automobile transportation systems. In this paper, a vehicular cloud computing-based SDIoV framework is studied wherein the joint allocation of transmission power and graph job is formulated as a nonlinear integer programming problem. To effectively address the problem, a structure-preservation-based two-stage allocation scheme is proposed that decouples template searching from power allocation. Specifically, a hierarchical tree-based r…
Dynamic nature of the cloud environment has made distributed resource management process a challenge for cloud service providers. The importance of maintaining the quality of service in accordance with customer expectations as well as the highly dynamic nature of cloud-hosted applications add new levels of complexity to the process. Advances to the big data learning approaches have shifted conventional static capacity planning solutions to complex performance-aware resource management methods. It is shown that the process of decision making for resource adjustment is closely related to the beh…
United Arab Emirates University（阿联酋大学）；College of Information Technology（信息技术学院）；National Water and Energy Center（国家水与能源中心）
Cloud computing enables remote execution of users tasks. The pervasive adoption of cloud computing in smart cities services and applications requires timely execution of tasks adhering to Quality of Services (QoS). However, the increasing use of computing servers exacerbates the issues of high energy consumption, operating costs, and environmental pollution. Maximizing the performance and minimizing the energy in a cloud data center is challenging. In this paper, we propose a performance and energy optimization bi-objective algorithm to tradeoff the contradicting performance and energy objecti…
Futurewei Technologies（华为技术有限公司）；Google, Inc.（谷歌公司）；Imperial College of London（伦敦帝国理工学院）；Lund University（隆德大学）；University of Wuertzburg（维尔茨堡大学）；University of Illinois at Urbana Champaign（伊利诺伊大学厄巴纳-香槟分校）；CMU Silicon Valley at the NASA Ames Research Center（卡内基梅隆大学硅谷校区NASA艾姆斯研究中心）；Delft University of Technology（代尔夫特理工大学）；INSA Lyon（里昂国立应用科学学院）
Elasticity in resource allocation is still a relevant problem in cloud computing. There are many academic and white papers which have investigated the problem and offered solutions.\textbf{Unfortunately, there are scant evidence of determining scaling number dynamically.} Elasticity is defined as the ability to adapt with the changing workloads by provisioning and de-provisioning Cloud resources. We propose ALVEC, a novel model of resource allocation in Cloud data centers, inspired by population dynamics and Mathematical Biology, which addresses dynamic allocation by auto-tuning model paramete…
With the pervasiveness of IoT devices, smart-phones and improvement of location-tracking technologies huge volume of heterogeneous geo-tagged (location specific) data is generated which facilitates several location-aware services. The analytics with this spatio-temporal (having location and time dimensions) datasets provide varied important services such as, smart transportation, emergency services (health-care, national defence or urban planning). While cloud paradigm is suitable for the capability of storage and computation, the major bottleneck is network connectivity loss. In time-critical…
