Split (1)

train · 127k rows

text large_stringlengths 7 3.02k
For large scale distributed storage systems, flash memories are an excellent choice because flash memories consume less power, take lesser floor space for a target throughput and provide faster access to data. In a traditional distributed filesystem, even distribution is required to ensure load-balancing, balanced spa...
We propose an efficient algorithm for the immersed boundary method on distributed-memory architectures, with the computational complexity of a completely explicit method and excellent parallel scaling. The algorithm utilizes the pseudo-compressibility method recently proposed by Guermond and Minev [Comptes Rendus Math...
We present BeRGeR: the first asynchronous geometric routing algorithm that guarantees delivery of a message despite a Byzantine fault without relying on cryptographic primitives or randomization. The communication graph is a planar embedding that remains three-connected if all edges intersecting the source-target line...
This paper proposes a novel exact distributed asynchronous subgradient-push algorithm (AsySPA) to solve an additive cost optimization problem over directed graphs where each node only has access to a local convex function and updates asynchronously with an arbitrary rate. Specifically, each node of a strongly connecte...
The computing continuum extends the high-performance cloud data centers with energy-efficient and low-latency devices close to the data sources located at the edge of the network. However, the heterogeneity of the computing continuum raises multiple challenges related to application management.
We have developed a highly scalable application, called Shoal, for tracking and utilizing a distributed set of HTTP web caches. Squid servers advertise their existence to the Shoal server via AMQP messaging by running Shoal Agent.
Graphics Processing Units (GPUs) have traditionally relied on the host CPU to initiate access to the data storage. This approach is well-suited for GPU applications with known data access patterns that enable partitioning of their dataset to be processed in a pipelined fashion in the GPU.
We describe a proposal for an extensible, component-based software architecture for natural language engineering applications. Our model leverages existing linguistic resource description and discovery mechanisms based on extended Dublin Core metadata.
Multireader shared registers are basic objects used as communication medium in asynchronous concurrent computation. We propose a surprisingly simple and natural scheme to obtain several wait-free constructions of bounded 1-writer multireader registers from atomic 1-writer 1-reader registers, that is easier to prove co...
This paper provides some new approaches to MPI implementations to improve MPI performance. These approaches include dynamically composable libraries, reducing average layer numbers of MPI libraries, and a single entity of MPI-network, MPI-protocol, and MPI.
The recent trend toward deep learning has led to the development of a variety of highly innovative AI accelerator architectures. One such architecture, the Cerebras Wafer-Scale Engine 2 (WSE-2), features 40 GB of on-chip SRAM, making it a potentially attractive platform for latency- or bandwidth-bound HPC simulation w...
We develop a simple and generic method to analyze randomized rumor spreading processes in fully connected networks. In contrast to all previous works, which heavily exploit the precise definition of the process under investigation, we only need to understand the probability and the covariance of the events that uninfo...
\textsc{Arbitrary Pattern Formation} is a fundamental problem in autonomous mobile robot systems. The problem asks to design a distributed algorithm that moves a team of autonomous, anonymous and identical mobile robots to form any arbitrary pattern $F$ given as input.
Many extreme scale scientific applications have workloads comprised of a large number of individual high-performance tasks. The Pilot abstraction decouples workload specification, resource management, and task execution via job placeholders and late-binding.
Convolution layers are prevalent in many classes of deep neural networks, including Convolutional Neural Networks (CNNs) which provide state-of-the-art results for tasks like image recognition, neural machine translation and speech recognition. The computationally expensive nature of a convolution operation has led to...
This study explores the integration of Agent AI with LangGraph to enhance real-time data analysis systems in big data environments. The proposed framework overcomes limitations of static workflows, inefficient stateful computations, and lack of human intervention by leveraging LangGraph's graph-based workflow cons...
We study online federated learning over a wireless network, where the central server updates an online global model sequence to minimize the time-varying loss of multiple local devices over time. The server updates the global model through over-the-air model-difference aggregation from the local devices over a noisy m...
Many HPC applications suffer from a bottleneck in the shared caches, instruction execution units, I/O or memory bandwidth, even though the remaining resources may be underutilized. It is hard for developers and runtime systems to ensure that all critical resources are fully exploited by a single application, so an att...
In this paper, we propose a paradigm for processing in parallel graph joins in road networks. The methodology we present can be used for distance join processing among the elements of two disjoint sets R,S of nodes from the road network, with R preceding S, and we are in search for the pairs of vertices (u,v), where u...
Edge computing has emerged as a key technology to reduce network traffic, improve user experience, and enable various Internet of Things applications. From the perspective of a service provider (SP), how to jointly optimize the service placement, sizing, and workload allocation decisions is an important and challengin...
We present a scheme to convert self-stabilizing algorithms that use randomization during and following convergence to self-stabilizing algorithms that use randomization only during convergence. We thus reduce the number of random bits from an infinite number to a bounded number.
Distributed frameworks are gaining increasingly widespread use in applications that process large amounts of data. One important example application is large scale similarity search, for which Locality Sensitive Hashing (LSH) has emerged as the method of choice, specially when the data is high-dimensional.
Recently, vision transformer (ViT) has started to outpace the conventional CNN in computer vision tasks. Considering privacy-preserving distributed learning with ViT, federated learning (FL) communicates models, which becomes ill-suited due to ViT' s large model size and computing costs.
We show that DNN accelerator micro-architectures and their program mappings represent specific choices of loop order and hardware parallelism for computing the seven nested loops of DNNs, which enables us to create a formal taxonomy of all existing dense DNN accelerators. Surprisingly, the loop transformations needed ...
Whereas deterministic protocols are typically guaranteed to obtain particular goals of interest, probabilistic protocols typically provide only probabilistic guarantees. This paper initiates an investigation of the interdependence between actions and subjective beliefs of agents in a probabilistic setting.
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.
Transient faults corrupt the content and organization of data structures. A recovery technique dealing with such faults is stabilization, which guarantees, following some number of operations on the data structure, that content of the data structure is legitimate.
Vectorization process focus on grouping pixels of a raster image into raw line segments, and forming lines, polylines or poligons. To vectorize massive raster images regarding resource and performane problems, weuse a distributed HIPI image processing interface based on MapReduce approach.
In the theoretical study of distributed communication networks, "history trees" are a discrete structure that naturally models the concept that anonymous agents become distinguishable upon receiving different sets of messages from neighboring agents. By conveniently organizing temporal information in a systema...
Lock-free data objects offer several advantages over their blocking counterparts, such as being immune to deadlocks and convoying and, more importantly, being highly concurrent. But they share a common disadvantage in that the operations they provide are difficult to compose into larger atomic operations while still g...
Transformer-based models are becoming deeper and larger recently. For better scalability, an underlying training solution in industry is to split billions of parameters (tensors) into many tasks and then run them across homogeneous accelerators (e.g., GPUs).
Swarms of autonomous devices are increasing in ubiquity and size. There are two main trains of thought for controlling devices in such swarms; centralized and distributed control.
Virtualization of distributed real-time systems enables the consolidation of mixed-criticality functions on a shared hardware platform thus easing system integration. Time-triggered communication and computation can act as an enabler of safe hard real-time systems.
We introduce a decentralized and online path planning technique for a network of unmanned aerial vehicles (UAVs) in the presence of weather disturbances. In our problem setting, the group of UAVs are required to collaboratively visit a set of goals scattered in a 2-dimensional area.
We uncover the extend-only directed posets (EDP) structure as a unification of recently discussed DAG-based Byzantine-tolerant conflict-free replicated data types (CRDT). We also show how a key-value map model can be derived from the EDP formulation, and give an outlook on an EDP-based systemic access control CRDT as ...
Initially, a number of frequent itemset mining (FIM) algorithms have been designed on the Hadoop MapReduce, a distributed big data processing framework. But, due to heavy disk I/O, MapReduce is found to be inefficient for such highly iterative algorithms.
GPUs are widely used to accelerate the training of machine learning workloads. As modern machine learning models become increasingly larger, they require a longer time to train, leading to higher GPU energy consumption.
Serverless computing has emerged as an attractive deployment option for cloud applications in recent times. The unique features of this computing model include, rapid auto-scaling, strong isolation, fine-grained billing options and access to a massive service ecosystem which autonomously handles resource management de...
We propose a dynamic verification framework for protocols in real-time distributed systems. The framework is based on Scribble, a tool-chain for design and verification of choreographies based on multiparty session types, developed with our industrial partners.
Evaluating how well a whole system or set of subsystems performs is one of the primary objectives of performance testing. We can tell via performance assessment if the architecture implementation meets the design objectives.
State-of-the-art asynchronous Byzantine fault-tolerant (BFT) protocols, such as HoneyBadgerBFT, BEAT, and Dumbo, have shown a performance comparable to partially synchronous BFT protocols. This paper studies two practical directions in asynchronous BFT.
Microservice architecture is widely adopted among distributed systems. It follows the modular approach that decomposes large software applications into independent services.
Multimodal Large Language Models (MLLMs) have been rapidly advancing, enabling cross-modal understanding and generation, and propelling artificial intelligence towards artificial general intelligence. However, existing MLLM inference systems are typically designed based on the architecture of language models, integrat...
High Performance Computing (HPC) systems are used across a wide range of disciplines for both large and complex computations. HPC systems often receive many thousands of computational tasks at a time, colloquially referred to as jobs.
Blockchain protocols typically aspire to run in the permissionless setting, in which nodes are owned and operated by a large number of diverse and unknown entities, with each node free to start or stop running the protocol at any time. This setting is more challenging than the traditional permissioned setting, in whic...
Molecular dynamics (MD) simulations are widely used to study large-scale molecular systems. HPC systems are ideal platforms to run these studies, however, reaching the necessary simulation timescale to detect rare processes is challenging, even with modern supercomputers.
Error-correcting codes are efficient methods for handling \emph{noisy} communication channels in the context of technological networks. However, such elaborate methods differ a lot from the unsophisticated way biological entities are supposed to communicate.
There has been considerable advancement in the use of blockchain for trust management in large-scale dynamic systems. In such systems, blockchain is mainly used to store the trust score or trust-related information of interactions among the various entities.
Federated Learning (FL) is a distributed machine learning (ML) paradigm, aiming to train a global model by exploiting the decentralized data across millions of edge devices. Compared with centralized learning, FL preserves the clients' privacy by refraining from explicitly downloading their data.
We consider unknown ad-hoc radio networks, when the underlying network is bidirectional and nodes can have polynomially large labels. For this model, we present a deterministic protocol for gossiping which takes $O(n \lg^2 n \lg \lg n)$ rounds.
We present a holistic design for GPU-accelerated computation in TrustZone TEE. Without pulling the complex GPU software stack into the TEE, we follow a simple approach: record the CPU/GPU interactions ahead of time, and replay the interactions in the TEE at run time.
Cloud-RAN is a recent architecture for mobile networks where the processing units are located in distant data centers while, until now, they were attached to antennas. The main challenge, to fulfill protocol constraints, is to guarantee low latency for the periodic messages sent from each antenna to its processing uni...
With a view to managing the increasing traffic in computer networks, round robin arbiter has been proposed to work with packet switching system to have increased speed in providing access and scheduling. Round robin arbiter is a doorway to a particular bus based on request along with equal priority and gives turns to ...
We demonstrate algorithm-based fault tolerance for silent, transient data corruption in "black-box" preconditioners. We consider both additive Schwarz domain decomposition with an ILU(k) subdomain solver, and algebraic multigrid, both implemented in the Trilinos library.
Increasing popularity of the serverless computing approach has led to the emergence of new cloud infrastructures working in Container-as-a-Service (CaaS) model like AWS Fargate, Google Cloud Run, or Azure Container Instances. They introduce an innovative approach to running cloud containers where developers are freed ...
Safety-critical systems must always have predictable and reliable behavior, otherwise systems fail and lives are put at risk. Even with the most rigorous testing it is impossible to test systems using all possible inputs.
Eventually linearizable objects are novel shared memory programming constructs introduced as an analogy to eventual consistency in message-passing systems. However, their behaviors in shared memory systems are so mysterious that very little general theoretical properties of them is known.
In the literature, there are several well-known periodic channel hopping (CH) sequences that can achieve maximum rendezvous diversity in a cognitive radio network (CRN). For a CRN with $N$ channels, it is known that the period of such a CH sequence is at least $N^2$.
An out-of-core stencil computation code handles large data whose size is beyond the capacity of GPU memory. Whereas, such an code requires streaming data to and from the GPU frequently.
Stencil computation is an important class of scientific applications that can be efficiently executed by graphics processing units (GPUs). Out-of-core approach helps run large scale stencil codes that process data with sizes larger than the limited capacity of GPU memory.
Achieving resource efficiency while preserving end-user experience is non-trivial for cloud application operators. As cloud applications progressively adopt microservices, resource managers are faced with two distinct levels of system behavior: end-to-end application latency and per-service resource usage.
In Decentralized Applications, off-chain storage solutions such as the InterPlanetary File System (IPFS) are crucial in overcoming Blockchain storage limitations. However, the assurance of data permanency in IPFS relies on the pinning of data, which comes with trust issues and potential single points of failure.
Today's high-performance computing (HPC) systems are heavily instrumented, generating logs containing information about abnormal events, such as critical conditions, faults, errors and failures, system resource utilization, and about the resource usage of user applications. These logs, once fully analyzed and corr...
We introduce a family of PCNs (Payment Channel Networks) characterized by a semi-hierarchical topology and a custom set of channel rebalancing strategies. This family exhibits two interesting benefits, if used as a platform for large-scale, instant, retail payment systems, such as CBDCs: Technically, the solution offe...
The Internet of Things (IoT) is emerging as the next big wave of digital presence for billions of devices on the Internet. Smart Cities are practical manifestation of IoT, with the goal of efficient, reliable and safe delivery of city utilities like water, power and transport to residents, through their intelligent ma...
In order to support the real-time interaction with LLMs and the instant search or the instant recommendation on social media, it becomes an imminent problem to build a k-NN graph or an indexing graph for the massive number of vectorized multimedia data. In such scenarios, the scale of the data or the scale of the grap...
Distributed training is an effective way to accelerate the training process of large-scale deep learning models. However, the parameter exchange and synchronization of distributed stochastic gradient descent introduce a large amount of communication overhead.
Modern concurrent programming benefits from a large variety of synchronization techniques. These include conventional pessimistic locking, as well as optimistic techniques based on conditional synchronization primitives or transactional memory.
Transparently checkpointing MPI for fault tolerance and load balancing is a long-standing problem in HPC. The problem has been complicated by the need to provide checkpoint-restart services for all combinations of an MPI implementation over all network interconnects.
Analyzing a distributed computation is a hard problem in general due to the combinatorial explosion in the size of the state-space with the number of processes in the system. By abstracting the computation, unnecessary explorations can be avoided.
We describe here a structured system for distributed mechanism design appropriate for both Intranet and Internet applications. In our approach the players dynamically form a network in which they know neither their neighbours nor the size of the network and interact to jointly take decisions.
We demonstrate the first-ever nontrivial, biologically realistic connectome simulated on neuromorphic computing hardware. Specifically, we implement the whole-brain connectome of the adult Drosophila melanogaster (fruit fly) from the FlyWire Consortium containing 140K neurons and 50M synapses on the Intel Loihi 2 neur...
Mobile edge computing (a.k.a. fog computing) has recently emerged to enable \emph{in-situ} processing of delay-sensitive applications at the edge of mobile networks. Providing grid power supply in support of mobile edge computing, however, is costly and even infeasible (in certain rugged or under-developed areas), thu...
Page placement is a critical problem for memoryintensive applications running on a shared-memory multiprocessor with a non-uniform memory access (NUMA) architecture. State-of-the-art page placement mechanisms interleave pages evenly across NUMA nodes.
Sparse linear algebra routines are fundamental building blocks of a large variety of scientific applications. Direct solvers, which are methods for solving linear systems via the factorization of matrices into products of triangular matrices, are commonly used in many contexts.
Training large-scale language models is increasingly critical in various domains, but it is hindered by frequent failures, leading to significant time and economic costs. Current failure recovery methods in cloud-based settings inadequately address the diverse and complex scenarios that arise, focusing narrowly on era...
It is commonly agreed that highly parallel software on Exascale computers will suffer from many more runtime failures due to the decreasing trend in the mean time to failures (MTTF). Therefore, it is not surprising that a lot of research is going on in the area of fault tolerance and fault mitigation.
Tensor cores (TCs) are a type of Application-Specific Integrated Circuit (ASIC) and are a recent addition to Graphics Processing Unit (GPU) architectures. As such, TCs are purposefully designed to greatly improve the performance of Matrix Multiply-Accumulate (MMA) operations.
Center-based clustering is a fundamental primitive for data analysis and becomes very challenging for large datasets. In this paper, we focus on the popular $k$-center variant which, given a set $S$ of points from some metric space and a parameter $k<\|S\|$, requires to identify a subset of $k$ centers in $S$ minimiz...
Coupled AI-Simulation workflows are becoming the major workloads for HPC facilities, and their increasing complexity necessitates new tools for performance analysis and prototyping of new in-situ workflows. We present SimAI-Bench, a tool designed to both prototype and evaluate these coupled workflows.
We consider the job assignment problem in a multi-server system consisting of $N$ parallel processor sharing servers, categorized into $M$ ($\ll N$) different types according to their processing capacity or speed. Jobs of random sizes arrive at the system according to a Poisson process with rate $N \lambda$.
Cloud providers sell their idle capacity on markets through an auction-like mechanism to increase their return on investment. The instances sold in this way are called spot instances.
In this paper we present a rudimentary model for low-latency anonymous communication systems. Specifically, we study distributed OR algorithm as an abstract of the system.
Communication scheduling aims to reduce communication bottlenecks in data parallel training (DP) by maximizing the overlap between computation and communication. However, existing schemes fall short due to three main issues: (1) hard data dependencies break some overlapping between communication and computation; (2) h...
The increasing scale and wealth of inter-connected data, such as those accrued by social network applications, demand the design of new techniques and platforms to efficiently derive actionable knowledge from large-scale graphs. However, real-world graphs are famously difficult to process efficiently.
Running multiple deep neural networks (DNNs) in parallel has become an emerging workload in both edge devices, such as mobile phones where multiple tasks serve a single user for daily activities, and data centers, where various requests are raised from millions of users, as seen with large language models. To reduce t...
Atomic Commit Problem (ACP) is a single-shot agreement problem similar to consensus, meant to model the properties of transaction commit protocols in fault-prone distributed systems. We argue that ACP is too restrictive to capture the complexities of modern transactional data stores, where commit protocols are integra...
Spherical Harmonic Transforms (SHT) are at the heart of many scientific and practical applications ranging from climate modelling to cosmological observations. In many of these areas new, cutting-edge science goals have been recently proposed requiring simulations and analyses of experimental or observational data at ...
Scientific applications that run on leadership computing facilities often face the challenge of being unable to fit leading science cases onto accelerator devices due to memory constraints (memory-bound applications). In this work, the authors studied one such US Department of Energy mission-critical condensed matter ...
Multiplication of two sparse matrices is a key operation in the simulation of the electronic structure of systems containing thousands of atoms and electrons. The highly optimized sparse linear algebra library DBCSR (Distributed Block Compressed Sparse Row) has been specifically designed to efficiently perform such sp...
Effective attention modules have played a crucial role in the success of Transformer-based large language models (LLMs), but the quadratic time and memory complexities of these attention modules also pose a challenge when processing long sequences. One potential solution for the long sequence problem is to utilize dis...
The Global-Local non-invasive coupling is an improvement of the submodeling technique, which permits to locally enhance structure computations by introducing patches with refined models and to take into accounts all the interactions. In order to circumvent its inherently limited computational performance, we propose a...
This paper proposes a new decentralized framework, named EDGChain-E (Encrypted-Data-Git Chain for Energy), designed to manage version-controlled, encrypted energy data using blockchain and the InterPlanetary File System. The framework incorporates a Decentralized Autonomous Organization (DAO) to orchestrate collaborat...
Function-as-a-Service is a popular cloud programming model that supports developers by abstracting away most operational concerns with automatic deployment and scaling of applications. Due to the high level of abstraction, developers rely on the cloud platform to offer a consistent service level, as decreased performa...
In recent years, large language models have achieved great success due to their unprecedented size. However, training these models poses a challenge for most researchers as it requires a substantial number of GPUs.
Byzantine agreement protocols in asynchronous networks have gained renewed attention due to their independence from network timing assumptions to ensure termination. Traditional asynchronous Byzantine agreement protocols require every party to broadcast its requests (e.g., transactions), leading to high communication ...
We study deterministic distributed broadcasting in synchronous multiple-access channels. Packets are injected into $n$ nodes by a window-type adversary that is constrained by a window $w$ and injection rates individually assigned to all nodes.
We study graph realization problems from a distributed perspective and we study it in the node capacitated clique (NCC) model of distributed computing, recently introduced for representing peer-to-peer networks. We focus on two central variants, degree-sequence realization and minimum threshold-connectivity realizatio...
Emerging technologies and applications including Internet of Things (IoT), social networking, and crowd-sourcing generate large amounts of data at the network edge. Machine learning models are often built from the collected data, to enable the detection, classification, and prediction of future events.
Most modern asset transfer systems use consensus to maintain a totally ordered chain of transactions. It was recently shown that consensus is not always necessary for implementing asset transfer.

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.