text large_stringlengths 7 3.02k |
|---|
Edge computing is a fast-growing computing paradigm where data is processed at the local site where it is generated, close to the end-devices. This can benefit a set of disruptive applications like autonomous driving, augmented reality, and collaborative machine learning, which produce incredible amounts of data that ... |
In this paper, we present a Byzantine fault tolerant distributed commit protocol for transactions running over untrusted networks. The traditional two-phase commit protocol is enhanced by replicating the coordinator and by running a Byzantine agreement algorithm among the coordinator replicas. |
In this paper we present two parallel Monte Carlo based algorithms for pricing multi--dimensional Bermudan/American options. First approach relies on computation of the optimal exercise boundary while the second relies on classification of continuation and exercise values. |
Function-as-a-Service (FaaS) allows to directly submit function code to a cloud provider without the burden of managing infrastructure resources. Each cloud provider establishes execution time limits to their FaaS offerings, which impose the risk of spending computation time without achieving partial results. |
Cloud monitoring activity involves dynamically tracking the Quality of Service (QoS) parameters related to virtualized resources (e.g., VM, storage, network, appliances, etc.), the physical resources they share, the applications running on them and data hosted on them. Applications and resources configuration in cloud... |
Computing on graphics processing units (GPUs) has become standard in scientific computing, allowing for incredible performance gains over classical CPUs for many computational methods. As GPUs were originally designed for 3D rendering, they still have several features for that purpose that are not used in scientific c... |
We consider parallel simulations for asynchronous systems employing L processing elements that are arranged on a ring. Processors communicate only among the nearest neighbors and advance their local simulated time only if it is guaranteed that this does not violate causality. |
Blockchains face a scalability challenge due to the intrinsic throughput limitations of consensus protocols and the limitation in block sizes due to decentralization. An alternative to improve the number of transactions per second is to use Layer 2 (L2) rollups. |
The aim of the dispersion problem is to place a set of $k(\leq n)$ mobile robots in the nodes of an unknown graph consisting of $n$ nodes such that in the final configuration each node contains at most one robot, starting from any arbitrary initial configuration of the robots on the graph. In this work we propose a va... |
In the ever evolving landscape of deep learning, unlocking the potential of cutting-edge models demands computational resources that surpass the capabilities of individual machines. Enter the NVIDIA DeepOps Slurm cluster, a meticulously orchestrated symphony of high-performance nodes, each equipped with powerful GPUs ... |
The Preconditioned Conjugate Gradient (PCG) method is widely used for solving linear systems of equations with sparse matrices. A recent version of PCG, Pipelined PCG, eliminates the dependencies in the computations of the PCG algorithm so that the non-dependent computations can be overlapped with communication. |
During the last decade or so, we have had a deluge of data from not only science fields but also industry and commerce fields. Although the amount of data available to us is constantly increasing, our ability to process it becomes more and more difficult. |
In this paper, we consider the problem of finding weak independent sets in a distributed network represented by a hypergraph. In this setting, each edge contains a set of r vertices rather than simply a pair, as in a standard graph. |
Stochastic simulation techniques are used for portfolio risk analysis. Risk portfolios may consist of thousands of reinsurance contracts covering millions of insured locations. |
When orchestrating highly distributed and data-intensive Web service workflows the geographical placement of the orchestration engine can greatly affect the overall performance of a workflow. Orchestration engines are typically run from within an organisations' network, and may have to transfer data across long ge... |
In this paper, we address the problem of broadcasting in a wireless network under a novel communication model: the {\em swamping} communication model. In this model, nodes communicate only with those nodes at geometric distance greater than $s$ and at most $r$ from them. |
Federated learning (FL) brings collaborative intelligence into industries without centralized training data to accelerate the process of Industry 4.0 on the edge computing level. FL solves the dilemma in which enterprises wish to make the use of data intelligence with security concerns. |
Training multi-billion to trillion-parameter language models efficiently on GPU clusters requires leveraging multiple parallelism strategies. We present Galvatron, a novel open-source framework (dubbed 'Optimus-Megatron' in the implementation) that dynamically combines data parallelism, tensor model parallelis... |
The Internet of Things is growing at a dramatic rate and extending into various application domains. We have designed, implemented and evaluated a resilient and decentralized system using IoT concepts. |
Serverless platforms such as Kubernetes are increasingly adopted in high-performance computing, yet autoscaling remains challenging under highly dynamic and heterogeneous workloads. Existing approaches often rely on uniform reactive policies or unconditioned predictive models, ignoring both workload semantics and pred... |
Function-as-a-Service (FaaS) is a popular cloud computing model in which applications are implemented as work flows of multiple independent functions. While cloud providers usually offer composition services for such workflows, they do not support cross-platform workflows forcing developers to hardcode the composition... |
A fundamental problem in distributed computing is the distribution of requests to a set of uniform servers without a centralized controller. Classically, such problems are modeled as static balls into bins processes, where $m$ balls (tasks) are to be distributed to $n$ bins (servers). |
Graphs are arguably one of the most fundamental data-structure used in many domains such as block-chain, networks etc. Theoretically and practically, improving Graph performance is one of the most studied and omnipresent research problems. |
Many companies rely on Cloud infrastructures for their computation, communication and data storage requirements. While Cloud services provide some benefits, e.g., replacing high upfront costs for an IT infrastructure with a pay-as-you-go model, they also introduce serious concerns that are notoriously difficult to add... |
Federated learning (FL) based on the centralized design faces both challenges regarding the trust issue and a single point of failure. To alleviate these issues, blockchain-aided decentralized FL (BDFL) introduces the decentralized network architecture into the FL training process, which can effectively overcome the d... |
In this paper, we present a framework for moving compute and data between processing elements in a distributed heterogeneous system. The implementation of the framework is based on the LLVM compiler toolchain combined with the UCX communication framework. |
The coffea framework provides a new approach to High-Energy Physics analysis, via columnar operations, that improves time-to-insight, scalability, portability, and reproducibility of analysis. It is implemented with the Python programming language, the scientific python package ecosystem, and commodity big data techno... |
Mobile edge computing (MEC) emerges as a promising solution for servicing delay-sensitive tasks at the edge network. A body of recent literature started to focus on cost-efficient service placement and request scheduling. |
Analyzing massive complex networks yields promising insights about our everyday lives. Building scalable algorithms to do so is a challenging task that requires a careful analysis and an extensive evaluation. |
Modern high performance computing clusters heavily rely on accelerators to overcome the limited compute power of CPUs. These supercomputers run various applications from different domains such as simulations, numerical applications or artificial intelligence (AI). |
Microservice and serverless computing systems open up massive versatility and opportunity to distributed and datacenter-scale computing. In the meantime, the deployments of modern datacenter resources are moving to disaggregated architectures. |
This paper focuses on the design and implementing of GPU-accelerated Adaptive Inverse Distance Weighting (AIDW) interpolation algorithm. The AIDW is an improved version of the standard IDW, which can adaptively determine the power parameter according to the spatial points distribution pattern and achieve more accurate... |
Distributed property testing in networks has been introduced by Brakerski and Patt-Shamir (2011), with the objective of detecting the presence of large dense sub-networks in a distributed manner. Recently, Censor-Hillel et al. (2016) have shown how to detect 3-cycles in a constant number of rounds by a distributed alg... |
State-of-the-art stream processing platforms make use of checkpointing to support fault tolerance, where a "checkpoint tuple" flows through the topology to all operators, indicating a checkpoint and triggering a checkpoint operation. The checkpoint will enable recovering from any kind of failure, be it as loca... |
This paper presents a novel approach to event-based power modelling for embedded platforms that do not have a Performance Monitoring Unit (PMU). The method involves complementing the target hardware platform, where the physical power data is measured, with another platform on which the CPU performance data, that is ne... |
Deep learning (DL) applications are increasingly being deployed on HPC systems, to leverage the massive parallelism and computing power of those systems for DL model training. While significant effort has been put to facilitate distributed training by DL frameworks, fault tolerance has been largely ignored. |
For the first time proposed: a method for representing the projections of a graph in computer memory and a description based on it of a quick search for shortest paths in unweighted dynamic graphs. The spatial complexity of the projection description does not exceed $(d + 1)\times n$ words, where $d$ is the diameter a... |
Disaggregated memory breaks the boundary of monolithic servers to enable memory provisioning on demand. Using network-attached memory to provide memory expansion for memory-intensive applications on compute nodes can improve the overall memory utilization on a cluster and reduce the total cost of ownership. |
The educational system manages extensive documentation and paperwork, which can lead to human errors and sometimes abuse or fraud, such as the falsification of diplomas, certificates or other credentials. In fact, in the last years, multiple cases of fraud have been detected, which have a significant cost to society, ... |
Detecting and handling network partitions is a fundamental requirement of distributed systems. Although existing partition detection methods in arbitrary graphs tolerate unreliable networks, they either assume that all nodes are correct or that a limited number of nodes might crash. |
This article will take you on a journey to the core of blockchains, their Byzantine consensus engine, where HotStuff emerged as a new algorithmic foundation for the classical Byzantine generals consensus problem. <br>The first part of the article underscores the theoretical advances HotStuff enabled, including several... |
The aim of this paper is to show how simple interaction mechanisms, inspired by chemical systems, can provide the basic tools to design and analyze a mathematical model for achieving consensus in wireless sensor networks, characterized by balanced directed graphs. The convergence and stability of the model are first p... |
This paper introduces FlowUnits, a novel programming and deployment model that extends the traditional dataflow paradigm to address the unique challenges of edge-to-cloud computing environments. While conventional dataflow systems offer significant advantages for large-scale data processing in homogeneous cloud settin... |
We investigate the problem of reaching majority agreement in a disconnected network. We obtain conditions under which such an agreement is certainly possible/impossible, and observe that these coincide in the ternary case. |
Operating a scalable and reliable server application, such as publish/subscribe (pub/sub) systems, requires tremendous development efforts and resources. The emerging serverless paradigm simplifies the development and deployment of highly available applications by delegating most operational concerns to the cloud prov... |
Earth observation analytics have the potential to serve many time-sensitive applications. However, due to limited bandwidth and duration of ground-satellite connections, it takes hours or even days to download and analyze data from existing Earth observation satellites, making real-time demands like timely disaster re... |
Data-sharing scientific collaborations have particular characteristics, potentially different from the current peer-to-peer environments. In this paper we advocate the benefits of exploiting emergent patterns in self-configuring networks specialized for scientific data-sharing collaborations. |
Generative Artificial Intelligence (GenAI) applies models and algorithms such as Large Language Model (LLM) and Foundation Model (FM) to generate new data. GenAI, as a promising approach, enables advanced capabilities in various applications, including text generation and image processing. |
Collaborative filtering (CF) has been proven to be one of the most effective techniques for recommendation. Among all CF approaches, SimpleX is the state-of-the-art method that adopts a novel loss function and a proper number of negative samples. |
Large multi-tenant production clusters often have to handle a variety of jobs and applications with a variety of complex resource usage characteristics. It is non-trivial and non-optimal to manually create placement rules for scheduling that would decide which applications should co-locate. |
Dynamic pricing is a promising strategy to address the challenges of smart charging, as traditional time-of-use (ToU) rates and stationary pricing (SP) do not dynamically react to changes in operating conditions, reducing revenue for charging station (CS) vendors and affecting grid stability. Previous studies evaluate... |
The classification of the most used load balancing algorithms in distributed systems (including cloud technology, cluster systems, grid systems) is described. Comparative analysis of types of the load balancing algorithms is conducted in accordance with the classification, the advantages and drawbags of each type of t... |
Distributed applications are commonly based on overlay networks interconnecting their sites so that they can exchange information. For these overlay networks to preserve their functionality, they should be able to recover from various problems like membership changes or faults. |
The growth of Large Language Models (LLMs) has necessitated large-scale distributed training. Highly optimized frameworks, however, still suffer significant losses in Model FLOPS utilization (often below 50%) due to large communication volumes. |
This paper investigates a swarm of autonomous mobile robots in the Euclidean plane, under the semi-synchronous ($\cal SSYNC$) scheduler. Each robot has a target function to determine a destination point from the robots' positions. |
In this paper I describe some results on the use of virtual processors technology for parallelize some SPMD computational programs in a cluster environment. The tested technology is the INTEL Hyper Threading on real processors, and the programs are MATLAB 6.5 Release 13 scripts for floating points computation. |
Theoreticians have studied distributed algorithms in the radio network model for close to three decades. A significant fraction of this work focuses on lower bounds for basic communication problems such as wake-up (symmetry breaking among an unknown set of nodes) and broadcast (message dissemination through an unknown... |
The majority of fault-tolerant distributed algorithms are designed assuming a nominal corruption model, in which at most a fraction $f_n$ of parties can be corrupted by the adversary. However, due to the infamous Sybil attack, nominal models are not sufficient to express the trust assumptions in open (i.e., permission... |
The latency issue of the cloud-centric IoT management system has motivated Fog and Edge Computing (FEC) architecture, which distributes the tasks from the cloud to the edge resources such as routers, switches or the IoT devices themselves. Specifically, mobile sensors of IoT system can also carry certain tasks for FEC... |
Serverless computing has grown in popularity in recent years, with an increasing number of applications being built on Functions-as-a-Service (FaaS) platforms. By default, FaaS platforms support retry-based fault tolerance, but this is insufficient for programs that modify shared state, as they can unwittingly persist... |
Decentralized applications rely on non-centralized technical infrastructures and coordination principles. Without trusted third parties, their execution is not controlled by entities exercising centralized coordination but is instead realized through technologies supporting distribution such as blockchains and serverl... |
Coded distributed computing introduced by Li et al. in 2015 is an efficient approach to trade computing power to reduce the communication load in general distributed computing frameworks such as MapReduce. In particular, Li et al. show that increasing the computation load in the Map phase by a factor of $r$ can create... |
With the recent development of technology, wireless sensor networks are becoming an important part of many applications such as health and medical applications, military applications, agriculture monitoring, home and office applications, environmental monitoring, etc. Knowing the location of a sensor is important, but... |
In this paper we present a novel algorithm developed for computing the QR factorisation of extremely ill-conditioned tall-and-skinny matrices on distributed memory systems. The algorithm is based on the communication-avoiding CholeskyQR2 algorithm and its block Gram-Schmidt variant. |
We have extended the Falkon lightweight task execution framework to make loosely coupled programming on petascale systems a practical and useful programming model. This work studies and measures the performance factors involved in applying this approach to enable the use of petascale systems by a broader user communit... |
We show that efficient simulations of the Kardar-Parisi-Zhang interface growth in 2 + 1 dimensions and of the 3-dimensional Kinetic Monte Carlo of thermally activated diffusion can be realized both on GPUs and modern CPUs. In this article we present results of different implementations on GPUs using CUDA and OpenCL an... |
Deep learning frameworks have been widely deployed on GPU servers for deep learning applications in both academia and industry. In training deep neural networks (DNNs), there are many standard processes or algorithms, such as convolution and stochastic gradient descent (SGD), but the running performance of different f... |
In this work, we examine the performance, energy efficiency and usability when using Python for developing HPC codes running on the GPU. We investigate the portability of performance and energy efficiency between CUDA and OpenCL; between GPU generations; and between low-end, mid-range and high-end GPUs. |
This paper initiates formal analysis of a simple, distributed algorithm for community detection on networks. We analyze an algorithm that we call \textsc{Max-LPA}, both in terms of its convergence time and in terms of the "quality" of the communities detected. |
Non-volatile memory (NVM), also known as persistent memory, is an emerging paradigm for memory that preserves its contents even after power loss. NVM is widely expected to become ubiquitous, and hardware architectures are already providing support for NVM programming. |
The data center of tomorrow is a data center made up of heterogeneous systems, which will run heterogeneous workloads. The systems will be located as close as possible to the data. |
The paper, based on authors' experience from several distributed systems integration projects, summarizes briefly practical designer's view on methodological requirements and overall system organization, including clues as to the organization of the application layer, use of operating system and preferred commu... |
On-device training has become an increasingly popular approach to machine learning, enabling models to be trained directly on mobile and edge devices. However, a major challenge in this area is the limited memory available on these devices, which can severely restrict the size and complexity of the models that can be ... |
Mobile edge computing facilitates users to offload computation tasks to edge servers for meeting their stringent delay requirements. Previous works mainly explore task offloading when system-side information is given (e.g., server processing speed, cellular data rate), or centralized offloading under system uncertaint... |
An effective atomic cross-chain swap protocol is introduced by Herlihy [Herlihy, 2018] as a distributed coordination protocol in order to exchange assets across multiple blockchains among multiple parties. An atomic cross-chain swap protocol guarantees; (1) if all parties conform to the protocol, then all assets are e... |
New directions in computing and algorithms has lead to some new applications that have tolerance to imprecision. Although, These applications are creating large volumes of data which exceeds the capability of today's computing systems. |
This article describes a method for creating applications for cluster computing systems using the parallel BSF skeleton based on the original BSF (Bulk Synchronous Farm) model of parallel computations developed by the author earlier. This model uses the master/slave paradigm. |
In this work we propose a highly optimized version of a simulated annealing (SA) algorithm adapted to the more recently developed Graphic Processor Units (GPUs). The programming has been carried out with CUDA toolkit, specially designed for Nvidia GPUs. |
This paper explores the problem of reaching approximate consensus in synchronous point-to-point networks, where each pair of nodes is able to communicate with each other directly and reliably. We consider the mobile Byzantine fault model proposed by Garay '94 -- in the model, an omniscient adversary can corrupt up... |
Big data applications involving the analysis of large datasets becomes a critical part of many emerging paradigms such as smart cities, social networks and modern security systems. Cloud computing has developed as a mainstream for hosting big data applications by its ability to provide the illusion of infinite resourc... |
Population protocols are a model of distributed computation intended for the study of networks of independent computing agents with dynamic communication structure. Each agent has a finite number of states, and communication opportunities occur nondeterministically, allowing the agents involved to change their states ... |
Matrix-vector multiplication forms the basis of many iterative solution algorithms and as such is an important algorithm also for hierarchical matrices. However, due to its low computational intensity, its performance is typically limited by the available memory bandwidth. |
Automatically tuning software configuration for optimizing a single performance attribute (e.g., minimizing latency) is not trivial, due to the nature of the configuration systems (e.g., complex landscape and expensive measurement). To deal with the problem, existing work has been focusing on developing various effect... |
Serverless computing has seen rapid adoption due to its high scalability and flexible, pay-as-you-go billing model. In serverless, developers structure their services as a collection of functions, sporadically invoked by various events like clicks. |
The Bulk-Synchronous Parallel model of computation has been used for the architecture independent design and analysis of parallel algorithms whose performance is expressed not only in terms of problem size n but also in terms of parallel machine properties. In this paper the performance of implementations of determini... |
The tremendous increase in the size and heterogeneity of supercomputers makes it very difficult to predict the performance of a scheduling algorithm. Therefore, dynamic solutions, where scheduling decisions are made at runtime have overpassed static allocation strategies. |
TensorFlow has been the most widely adopted Machine/Deep Learning framework. However, little exists in the literature that provides a thorough understanding of the capabilities which TensorFlow offers for the distributed training of large ML/DL models that need computation and communication at scale. |
Payment channel networks (PCNs) are a promising technology that alleviates blockchain scalability by shifting the transaction load from the blockchain to the PCN. Nevertheless, the network topology has to be carefully designed to maximise the transaction throughput in PCNs. |
Given the on-demand nature of cloud computing, managing cloud-based services requires accurate modeling for the correlation between their Quality of Service (QoS) and cloud configurations/resources. The resulted models need to cope with the dynamic fluctuation of QoS sensitivity and interference. |
Improving transaction throughput is an important challenge for Bitcoin. However, shortening the block generation interval or increasing the block size to improve throughput makes it sharing blocks within the network slower and increases the number of orphan blocks. |
Mobile Edge Computing (MEC) pushes computing functionalities away from the centralized cloud to the network edge, thereby meeting the latency requirements of many emerging mobile applications and saving backhaul network bandwidth. Although many existing works have studied computation offloading policies, service cachi... |
Convolutional Neural Networks (CNNs) are rapidly gaining popularity in varied fields. Due to their increasingly deep and computationally heavy structures, it is difficult to deploy them on energy constrained mobile applications. |
A local algorithm is a distributed algorithm that completes after a constant number of synchronous communication rounds. We present local approximation algorithms for the minimum dominating set problem and the maximum matching problem in 2-coloured and weakly 2-coloured graphs. |
This paper presents GearV, a two-gear lightweight hypervisor architecture to address the some known challenges. By dividing hypervisor into some partitions, and dividing scheduling policies into Gear1 and Gear2 respectively, GearV creates a consolidated platform to run best-effort system and safety-critical system sim... |
This paper proposes a framework for distributed, in-storage training of neural networks on clusters of computational storage devices. Such devices not only contain hardware accelerators but also eliminate data movement between the host and storage, resulting in both improved performance and power savings. |
Broadcast is a ubiquitous distributed computing problem that underpins many other system tasks. In static, connected networks, it was recently shown that broadcast is solvable without any node memory and only constant-size messages in worst-case asymptotically optimal time (Hussak and Trehan, PODC'19/STACS'20/... |
Application-layer multicast implements the multicast functionality at the application layer. The main goal of application-layer multicast is to construct and maintain efficient distribution structures between end-hosts. |
In this paper we build a case for providing job completion time predictions to cloud users, similar to the delivery date of a package or arrival time of a booked ride. Our analysis reveals that providing predictability can come at the expense of performance and fairness. |
Cloud applications are increasingly shifting from large monolithic services, to complex graphs of loosely-coupled microservices. Despite their advantages, microservices also introduce cascading QoS violations in cloud applications, which are difficult to diagnose and correct. |
Deep learning (DL) is becoming increasingly popular in several application domains and has made several new application features involving computer vision, speech recognition and synthesis, self-driving automobiles, drug design, etc. feasible and accurate. As a result, large scale on-premise and cloud-hosted deep lear... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.