检索结果-内蒙古大学图书馆

17th USENIX symposium on Operating Systems Design and Implementation (OSDI)

作者： Wei, Xingda Lu, Fangming Wang, Tianxia Gu, Jinyu Yang, Yuhan Chen, Rong Chen, Haibo Shanghai Jiao Tong Univ Inst Parallel & Distributed Syst SEIEE Shanghai Peoples R China Shanghai AI Lab Shanghai Peoples R China

ISBN: (纸本)9781939133342

Serverless platforms essentially face a tradeoff between container startup time and provisioned concurrency (i.e., cached instances), which is further exaggerated by the frequent need for remote container initialization. This paper presents MITOSIS, an operating system primitive that provides fast remote fork, which exploits a deep codesign of the OS kernel with RDMA. By leveraging the fast remote read capability of RDMA and partial state transfer across serverless containers, MITOSIS bridges the performance gap between local and remote container initialization. MITOSIS is the first to fork over 10,000 new containers from one instance across multiple machines within a second, while allowing the new containers to efficiently transfer the pre-materialized states of the forked one. We have implemented MITOSIS on Linux and integrated it with FN, a popular serverless platform. Under load spikes in real-world serverless workloads, MITOSIS reduces the function tail latency by 89% with orders of magnitude lower memory usage. For serverless workflow that requires state transfer, MITOSIS improves its execution time by 86%.

关键词： Containers

来源：评论

学校读者我要写书评

暂无评论

distributed (Δ+1)-COLORING VIA ULTRAFAST GRAPH SHATTERING

引用

SIAM JOURNAL ON computing 2020年第3期49卷 497-539页

作者： Chang, Yi-Jun Li, Wenzheng Pettie, Seth Swiss Fed Inst Technol Inst Theoret Studies CH-8092 Zurich Switzerland Stanford Univ Comp Sci Dept Stanford CA 94305 USA Univ Michigan EECS Ann Arbor MI 48109 USA

Vertex coloring is one of the classic symmetry breaking problems studied in distributed computing. In this paper we present a new algorithm for (Delta+1)-list coloring in the randomized LOCAL model running in O(Det(d)(polylogn)) time, where Det(d)(n') is the deterministic complexity of (deg+1)-list coloring on n'-vertex graphs. (In this problem, each v has a palette of size deg(v)+1.) This improves upon a previous randomized algorithm of Harris, Schneider, and Su [J. acm, 65 (2018), 19] with complexity O(root log Delta+log log n+Det(d)(poly log n)) = O(root log n). Unless Delta is small, it is also faster than the best known deterministic algorithm of Fraigniaud, Heinrich, and Kosowski [proceedings of the 57th Annual IEEE symposium on Foundations of Computer Science (FOCS), 2016] and Barenboim, Elkin, and Goldenberg [proceedings of the 38th Annual acm symposium on principles of distributed computing (PODC), 2018], with complexity O(root Delta log log* Delta + log* n). Our algorithm's running time is syntactically very similar to the Omega(Det(poly log n)) lower bound of Chang, Kopelowitz, and Pettie [SIAM J. Comput., 48 (2019), pp. 122-143], where Det(n') is the deterministic complexity of (Delta + 1)-list coloring on n'-vertex graphs. Although distributed coloring has been actively investigated for 30 years, the best determin- istic algorithms for (deg +1)- and (Delta + 1)-list coloring (that depend on n' but not Delta) use a black-box application of network decompositions. The recent deterministic network decomposition algorithm of Rozhoil and Ghaffari [proceedings of the 52nd Annual acm SIGACT symposium on Theory of computing (STOC), 2020] implies that Det(d)(n') and Det(n') are both poly(log n'). Whether they are asymptotically equal is an open problem.

关键词： distributed algorithm local model graph coloring

来源：评论

学校读者我要写书评

暂无评论

3-Majority and 2-Choices with Many Opinions 25

3-Majority and 2-Choices with Many Opinions

引用

proceedings of the acm symposium on principles of distributed computing

作者： Nobutaka Shimizu Takeharu Shiraga Institute of Science Tokyo Tokyo Japan Chuo University Tokyo Japan

ISBN: (纸本)9798400718854

We present the first nearly-optimal bounds on the consensus time for the well-known synchronous consensus dynamics, specifically 3-Majority and 2-Choices, for an arbitrary number of opinions. In synchronous consensus dynamics, we consider an n-vertex complete graph with self-loops, where each vertex holds an opinion from {1,..., k}. At each discrete-time round, all vertices update their opinions simultaneously according to a given protocol. The goal is to reach a consensus, where all vertices support the same opinion. In 3-Majority, each vertex chooses three random neighbors with replacement and updates its opinion to match the majority, with ties broken randomly. In 2-Choices, each vertex chooses two random neighbors with replacement. If the selected vertices hold the same opinion, the vertex adopts that opinion. Otherwise, it retains its current opinion for that *** upon a line of work [Becchetti et al., SPAA'14], [Becchetti et al., SODA'16], [Berenbrink et al., PODC'17], [Ghaffari and Lengler, PODC'18], we prove that, for every 2 ≤ k ≤ n, 3-Majority (resp. 2-Choices) reaches consensus within [EQUATION] (resp. [EQUATION] rounds with high probability. Prior to this work, the best known upper bound on the consensus time of 3-Majority was Õ(k) if k ≪ n1/3 and Õ(n2/3) otherwise, and for 2-Choices, the consensus time was known to be Õ(k) for [EQUATION].

关键词： distributed voting

来源：评论

学校读者我要写书评

暂无评论

StencilFlow: Mapping Large Stencil Programs to distributed Spatial computing Systems 21

StencilFlow: Mapping Large Stencil Programs to Distributed S...

引用

19th IEEE/acm International symposium on Code Generation and Optimization (CGO)

作者： Licht, Johannes de Fine Kuster, Andreas De Matteis, Tiziano Ben-Nun, Tal Hofer, Dominic Hoefler, Torsten Swiss Fed Inst Technol Dept Comp Sci Zurich Switzerland MeteoSwiss Zurich Switzerland

ISBN: (纸本)9781728186139

Spatial computing devices have been shown to significantly accelerate stencil computations, but have so far relied on unrolling the iterative dimension of a single stencil operation to increase temporal locality. This work considers the general case of mapping directed acyclic graphs of heterogeneous stencil computations to spatial computing systems, assuming large input programs without an iterative component. StencilFlow maximizes temporal locality and ensures deadlock freedom in this setting, providing end-to-end analysis and mapping from a high-level program description to distributed hardware. We evaluate our generated architectures on a Stratix 10 FPGA testbed, yielding 1.31 TOp /s and 4.18 TOp/s on single-device and multi-device, respectively, demonstrating the highest performance recorded for stencil programs on FPGAs to date. We then leverage the framework to study a complex stencil program from a production weather simulation application. Our work enables productively targeting distributed spatial computing systems with large stencil programs, and offers insight into architecture characteristics required for their efficient execution in practice.

关键词： Mapping

来源：评论

学校读者我要写书评

暂无评论

Enabling Edge computing over LoRaWAN: A Device-Gateway Coordination Protocol 12

Enabling Edge Computing over LoRaWAN: A Device-Gateway Coord...

引用

12th acm International symposium on Design and Analysis of Intelligent Vehicular Networks and Applications, DIVANet 2022

作者： Fardin, Ivan Milani, Stefano Cuomo, Francesca Chatzigiannakis, Ioannis DIAG Sapienza University of Rome Rome Italy DIET Sapienza University of Rome Rome Italy CNIT Consorzio Nazionale Interuniversitario per le Telecomunicazioni Parma Italy

ISBN: (纸本)9781450394826

The freedom of the LoRaWAN license-free ad-hoc deployment model can significantly reduce the complexity of network deployment, however, it introduces certain problems that hinder network scalability and performance. First, Network Gateways simply forward frames to the centralised Network Server without the possibility of becoming Edge computing processing elements. Second, Network Gateways may have overlapping areas of network coverage (a) resulting in an increase of network traffic at the network backbone as frames are relayed to the Network Server multiple times, (b) may cause unexpected frame collisions and duty-cycle exhaustion. In this paper, a novel decentralised algorithm is presented that assigns each IoT Device to a single Network Gateway so that (a) duplicate message deliveries are completely avoided, (b) Network Gateways can become an intermediate operations layer between the IoT devices and the Network Server providing computational and storage resources. The proposed protocol is implemented in the OMNeT++ simulator and evaluated based on datasets collected from long-term real-world deployments. To improve the accuracy of the experiments in large-scale and dense urban deployments the OMNeT++ LoRa interference model is extended by considering the non-perfect orthogonality of the LoRa Spreading Factors and the channel adjacency. The performance evaluation highlights the benefits of the distributed approach proposed here and provides valuable indications on the overall performance of the network. © 2022 acm.

关键词： Scalability

来源：评论

学校读者我要写书评

暂无评论

SchedInspector: A Batch Job Scheduling Inspector Using Reinforcement Learning 22

SchedInspector: A Batch Job Scheduling Inspector Using Reinf...

引用

31st International symposium on High-Performance Parallel and distributed computing (HPDC)

作者： Zhang, Di Dai, Dong Xie, Bing Univ North Carolina Charlotte Comp Sci Dept Charlotte NC 28223 USA Oak Ridge Natl Lab Oak Ridge Leadership Comp Facil Oak Ridge TN USA

ISBN: (纸本)9781450391993

Improving the performance of job executions is an important goal of HPC batch job schedulers, such as minimizing job waiting time, slowdown, or completion time. Such a goal is often accomplished using carefully designed heuristics based on job features, such as job size and job duration. However, these heuristics overlook important runtime factors (e.g., cluster availability and waiting job patterns), which may vary across time and make a previously sound scheduling decision not hold any longer. In this study, we propose a new approach to incorporate runtime factors into batch job scheduling for better job execution performance. The key idea is to add a scheduling inspector on top of the base job scheduler to scrutinize its scheduling decisions. The inspector will take the runtime factors into consideration and accordingly determine the fitness of the scheduled job. It then either accepts the scheduled job or rejects it and asks the base schedulers to try again later. We realize such an inspector, namely SchedInspector, by leveraging the intelligence of reinforcement learning. Through extensive experiments, we show SchedInspector can intelligently integrate the runtime factors into various batch job scheduling policies, including the state-of-the-art one, to gain better job execution performance, such as smaller average bounded job slowdown (up to 69% better) or average job waiting time (up to 52% better), across various real-world workloads. We also show that although rejecting scheduling decisions may leave the resources idle hence affect the system utilization, SchedInspector is able to achieve the job execution performance improvement with marginal impact on the system utilization (typically less than 1%). We consider one key advantage of SchedInspector is it automatically learns to work with and improve existing job scheduling policies without changing them, which makes it promising to serve as a generic enhancer for various batch job scheduling policies.

关键词： High Performance computing (HPC) batch job scheduling reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Brief Announcement: New distributed Interactive Proofs for Planarity: A Matter of Left and Right 25

Brief Announcement: New Distributed Interactive Proofs for P...

引用

proceedings of the acm symposium on principles of distributed computing

作者： Yuval Gil Merav Parter Weizmann Institute of Science Rehovot Israel

ISBN: (纸本)9798400718854

We provide new distributed interactive proofs (DIP) for planarity and related graph families. The notion of a distributed interactive proof (DIP) was introduced by Kol, Oshman, and Saxena (PODC 2018). In this setting, the verifier consists of n nodes connected by a communication graph G. The prover is a single entity that communicates with all nodes by short messages. The goal is to verify that the graph G satisfies a certain property (e.g., planarity) in a small number of rounds, and with a small communication bound, denoted as the proof *** work by Naor, Parter and Yogev (SODA 2020) presented a DIP for planarity that uses three interaction rounds and a proof size of O (log n). Feuilloley et al. (PODC 2020) showed that the same can be achieved with a single interaction round and without randomization, by providing a proof labeling scheme with a proof size of O (log n). In this work, we design new DIPs that use exponentially shorter proofs compared to the state-of-the-art bounds. Our main results are:• There is a 5-round protocol with O (log log n) proof size for outerplanarity.• There is a 5-round protocol with O (log log n) proof size for verifying embedded planarity and O (log log n + log Δ) proof size for general planar graphs, where Δ is the maximum degree in the graph. In the former setting, it is assumed that an embedding of the graph is given (e.g., each node holds a clockwise orientation of its neighbors) and the goal is to verify that it is a valid planar embedding. The latter result should be compared with the non-interactive setting for which there is lower bound of Ω (log n) bits for graphs with Δ = O(1) by Feuilloley et al. (PODC 2020).• The non-interactive deterministic lower bound of Ω (log n) bits by Feuilloley et al. (PODC 2020) can be extended to hold even if the verifier is randomized. Moreover, the lower bound holds even with the assumption that the verifier's randomness comes in the form of an unbounded random string shared among the nod

关键词： distributed interactive proofs

来源：评论

学校读者我要写书评

暂无评论

Poster: Extending VLANs to the Client - Opportunistic Edge computing with EVPNs 23

Poster: Extending VLANs to the Client - Opportunistic Edge C...

引用

proceedings of the Eighth acm/IEEE symposium on Edge computing

作者： Stephan Recker Computer Science Fachhochschule Dortmund Dortmund Germany

ISBN: (纸本)9798400701238

Opportunistic Edge computing forms a pool of resources for delivery of edge computing services leveraging existing nearby infrastructure. The pooled resources are generally owned by different entities and unreliable. Workload hosted on such resources must be movable in case of infrastructure down-times. Additionally, distributed applications are requiring a one-to-many connectivity model rather than client-to-server connectivity. Therefore, the underlay transport network needs to cater for these requirements. This article proposes a concept that meets these based on standard EVPN/VXLAN tunnels. In order to avoid additional latency resulting from NAT and tromboning, EVPNs will need to be extended to the client. Besides introducing the networking concept also a critical assessment of resulting constraints for clients will be provided.

关键词： edge computing

来源：评论

学校读者我要写书评

暂无评论

FlexScience'22: 12th Workshop on AI and Scientific computing at Scale using Flexible computing Infrastructures 22

FlexScience'22: 12th Workshop on AI and Scientific Computing...

引用

31st International symposium on High-Performance Parallel and distributed computing (HPDC)

作者： Costan, Alexandru Nicolae, Bogdan Sato, Kento INSA Rennes IRISA Rennes France Argonne Natl Lab Lemont IL USA RIKEN Ctr Kobe Hyogo Japan

ISBN: (纸本)9781450391993

Scientific computing applications generate enormous datasets that are continuously increasing exponentially in both complexity and volume, making their analysis, archival, and sharing one of the grand challenges of modern big data analytics. Supported by the rise of artificial intelligence and deep learning, such enormous datasets are becoming valuable resources even beyond their original scope, opening new opportunities to learn patterns and extract new knowledge at large scale, potentially without human intervention. However, this leads to an increasing complexity of the workflows that combine traditional HPC simulations with big data analytics and AI applications. An initial wave that opened this direction was the shift from compute-intensive to data-intensive, which saw several ideas from big data analytics (in-situ processing, shipping computations close to data, complex and dynamic workflows) fused with the tightly coupled patterns addressed by the AI and the high performance computing ecosystems. In a quest to keep up with the complexity of the workflows, the design and operation of the infrastructures capable of running them efficiently at scale has evolved accordingly. Extreme heterogeneity at all levels (combinations of CPUs and accelerators, various types of memories and local storage and network links, parallel file systems and object stores, etc.) is now the norm. ideas pioneered by cloud and edge computing (aspects related to elasticity, multi-tenancy, geo-distributed processing, stream computing) are also beginning to be adopted in the HPC ecosystem (containerized workflows, on-demand jobs to complement batch jobs, streaming of experimental data from instruments directly to supercomputers, etc.). Thus, modern scientific applications need to be integrated into an entire Compute Continuum from the edge all the way to supercomputers and large data-centers using flexible infrastructures and middlewares. The 12th workshop on AI and Scientific computing at

关键词： computing continuum scientific computing cloud computing edge processing artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

The Cerebras CS-2: Designing an AI Accelerator around the World's Largest 2.6 Trillion Transistor Chip 22

The Cerebras CS-2: Designing an AI Accelerator around the Wo...

引用

31st edition of the acm International symposium on Physical Design (ISPD)

作者： Fricker, Jean-Philippe Cerebras Syst Sunnyvale CA 94085 USA

ISBN: (纸本)9781450392105

The computing and memory demands from state-of-the-art neural networks have increased several orders of magnitude in just the last couple of years, and there's no end in sight. Traditional forms of scaling chip performance are necessary but far from sufficient to run the machine learning models of the future. In this talk, Cerebras Co-Founder and Chief Systems Architect Jean-Philippe Fricker will explore the fundamental properties of neural networks and why they are not well served by traditional architectures. He will examine how co-design can relax the traditional boundaries between technologies and enable designs specialized for neural networks with new architectural capabilities and performance. Finally, Jean-Philippe will explore this rich new design space using the Cerebras architecture as a case study, highlighting design principles and tradeoffs that enable the machine learning models of the future.

关键词： computing Neural networks AI Wafer scale engine Clusterscale performance

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：