Related papers: Parallel Deferred Update Replication
The aim of this paper is to present a first evaluation of the potential of an asynchronous distributed computation associated to the recently proposed approach, D-iteration: the D-iteration is a fluid diffusion based iterative method, which…
Read-Copy Update (RCU) is a scalable, high-performance Linux-kernel synchronization mechanism that runs low-overhead readers concurrently with updaters. Production-quality RCU implementations for multi-core systems are decidedly…
In this paper we present two strategies to enable "parallelization across the method" for spectral deferred corrections (SDC). Using standard low-order time-stepping methods in an iterative fashion, SDC can be seen as preconditioned Picard…
Deep reinforcement learning (DRL) algorithms and evolution strategies (ES) have been applied to various tasks, showing excellent performances. These have the opposite properties, with DRL having good sample efficiency and poor stability,…
While distributed training significantly speeds up the training process of the deep neural network (DNN), the utilization of the cluster is relatively low due to the time-consuming data synchronizing between workers. To alleviate this…
Communication scheduling has been shown to be effective in accelerating distributed training, which enables all-reduce communications to be overlapped with backpropagation computations. This has been commonly adopted in popular distributed…
This paper considers the scheduling of parallel real-time tasks with arbitrary-deadlines. Each job of a parallel task is described as a directed acyclic graph (DAG). In contrast to prior work in this area, where decomposition-based…
Training large deep learning models requires parallelization techniques to scale. In existing methods such as Data Parallelism or ZeRO-DP, micro-batches of data are processed in parallel, which creates two drawbacks: the total memory…
Self-adjusting computation is an approach for automatically producing dynamic algorithms from static ones. The approach works by tracking control and data dependencies, and propagating changes through the dependencies when making an update.…
Conventional blockchains use consensus algorithms that totally order updates across all accounts, which is stronger than necessary to implement a replicated ledger. This makes updates slower and more expensive than necessary. More recent…
In this paper we present the new DUNE-ALUGrid module. This module contains a major overhaul of the sources from the ALUgrid library and the binding to the DUNE software framework. The main changes include user defined load balancing,…
This paper presents PIPQ, a strict and linearizable concurrent priority queue whose design differs from existing solutions in literature because it focuses on enabling parallelism of insert operations as opposed to accelerating delete-min…
A dominant cost for query evaluation in modern massively distributed systems is the number of communication rounds. For this reason, there is a growing interest in single-round multiway join algorithms where data is first reshuffled over…
Data is replicated and stored redundantly over multiple servers for availability in distributed databases. We focus on databases with frequent reads and writes, where both read and write latencies are important. This is in contrast to…
In order to increase availability in a distributed system some or all of the data items are replicated and stored at separate sites. This is an issue of key concern especially since there is such a proliferation of wireless technologies and…
One key requirement for storage clouds is to be able to retrieve data quickly. Recent system measurements have shown that the data retrieving delay in storage clouds is highly variable, which may result in a long latency tail. One crucial…
This paper offers a new hybrid probably approximately correct (PAC) reinforcement learning (RL) algorithm for Markov decision processes (MDPs) that intelligently maintains favorable features of its parents. The designed algorithm, referred…
Deep Neural Network (DNN) models are usually trained sequentially from one layer to another, which causes forward, backward and update locking's problems, leading to poor performance in terms of training time. The existing parallel…
Parallel real-time embedded applications can be modelled as directed acyclic graphs (DAGs) whose nodes model subtasks and whose edges model precedence constraints among subtasks. Efficiently scheduling such parallel tasks can be challenging…
Data movement between the processor and the main memory is a first-order obstacle against improving performance and energy efficiency in modern systems. To address this obstacle, Processing-using-Memory (PuM) is a promising approach where…