Task graphs have been studied for decades as a foundation for scheduling irregular parallel applications and incorporated in programming models such as OpenMP. While many high-performance parallel libraries are based on task graphs, they also have additional scheduling requirements, such as synchronization from inner levels of data parallelism and internal blocking communications. In this paper, we extend task-graph scheduling to support efficient synchronization and communication within tasks. Our scheduler avoids deadlock and oversubscription of worker threads, and refines victim selection to increase the overlap of sibling tasks. To the best of our knowledge, our approach is the first to combine gang-scheduling and work-stealing in a single runtime. Our approach has been evaluated on the SLATE highperformance linear algebra library. Relative to the LLVM OMP runtime, our runtime demonstrates performance improvements of up to 13.82%, 15.2%, and 36.94% for LU, QR, and Cholesky, respectively, evaluated across different configurations.
Work Stealing has been a very successful algorithm for scheduling parallel computations, and is known to achieve high performances even for computations exhibiting fine-grained parallelism. We present a variant of \ws\ that provably avoids most synchronization overheads by keeping processors' deques entirely private by default, and only exposing work when requested by thieves. This is the first paper that obtains bounds on the synchronization overheads that are (essentially) independent of the total amount of work, thus corresponding to a great improvement, in both algorithm design and theory,…
Most methods for Byzantine fault tolerance (BFT) in the partial synchrony setting divide the local state of the nodes into views, and the transition from one view to the next dictates a leader change. In order to provide liveness, all honest nodes need to stay in the same view for a sufficiently long time. This requires \emph{view synchronization}, a requisite of BFT that we extract and formally define here. Existing approaches for Byzantine view synchronization incur quadratic communication (in $n$, the number of parties). A cascade of $O(n)$ view changes may thus result in $O(n^3)$ communica…
For tolerating Byzantine faults of both the terminal and communication components in self-stabilizing clock synchronization, the two-dimensional self-stabilizing Byzantine-fault-tolerant clock synchronization problem is investigated and solved. By utilizing the time-triggered (TT) stage provided in the underlying networks as TT communication windows, the approximate agreement, hopping procedure, and randomized grandmasters are integrated into the overall solution. It is shown that with partitioning the communication components into $3$ arbitrarily connected subnetworks, efficient synchronizati…
The complexity of embedded application design is increasing with growing user demands. In particular, automotive embedded systems are highly complex in nature, and their functionality is realized by a set of periodic tasks. These tasks may have hard real-time requirements and communicate over an interconnect. The problem is to efficiently co-schedule task execution on cores and message transmission on the interconnect so that timing constraints are satisfied. Contemporary works typically deal with zero-jitter scheduling, which results in lower resource utilization, but has lower memory require…
Scheduling precedence-constrained tasks is a classical problem that has been studied for more than fifty years. However, little progress has been made in the setting where there are communication delays between tasks. Results for the case of identical machines were derived nearly thirty years ago, and yet no results for related machines have followed. In this work, we propose a new scheduler, Generalized Earliest Time First (GETF), and provide the first provable, worst-case approximation guarantees for the goals of minimizing both the makespan and total weighted completion time of tasks with p…
Model-based development is a widely-used method to describe complex systems that enables the rapid prototyping. Advances in the science of distributed systems has led to the development of large scale statechart models which are distributed among multiple locations. Taking medicine for example, models of best-practice guidelines during rural ambulance transport are distributed across hospital settings from a rural hospital, to an ambulance, to a central tertiary hospital. Unfortunately, these medical models require continuous and real-time communication across individual medical models in phys…
Understanding the bottlenecks in implementing stochastic gradient descent (SGD)-based distributed support vector machines (SVM) algorithm is important in training larger data sets. The communication time to do the model synchronization across the parallel processes is the main bottleneck that causes inefficiency in the training process. The model synchronization is directly affected by the mini-batch size of data processed before the global synchronization. In producing an efficient distributed model, the communication time in training model synchronization has to be as minimum as possible whi…
Reconfigurable broadcast networks provide a convenient formalism for modelling and reasoning about networks of mobile agents broadcasting messages to other agents following some (evolving) communication topology. The parameterized verification of such models aims at checking whether a given property holds irrespective of the initial configuration (number of agents, initial states and initial communication topology). We focus here on the synchronization property, asking whether all agents converge to a set of target states after some execution. This problem is known to be decidable in polynomia…
Department of Computer Science, Hong Kong Baptist University（香港浸会大学计算机科学系）
Distributed Deep Learning (DDL) has rapidly grown its popularity since it helps boost the training performance on high-performance GPU clusters. Efficient job scheduling is indispensable to maximize the overall performance of the cluster when training multiple jobs simultaneously. However, existing schedulers do not consider the communication contention of multiple communication tasks from different distributed training jobs, which could deteriorate the system performance and prolong the job completion time. In this paper, we first establish a new DDL job scheduling framework which organizes D…
We reduce the cost of communication and synchronization in graph processing by analyzing the fastest way to process graphs: pushing the updates to a shared state or pulling the updates to a private state.We investigate the applicability of this push-pull dichotomy to various algorithms and its impact on complexity, performance, and the amount of used locks, atomics, and reads/writes. We consider 11 graph algorithms, 3 programming models, 2 graph abstractions, and various families of graphs. The conducted analysis illustrates surprising differences between push and pull variants of different al…
Shenzhen Research Institute of Big Data（深圳大数据研究院）；Southern University of Science and Technology（南方科技大学）；China Academy of Information and Communications Technology（中国信息通信研究院）；The Chinese University of Hong Kong (Shenzhen)（香港中文大学（深圳））；Zhejiang University（浙江大学）
The popular federated edge learning (FEEL) framework allows privacy-preserving collaborative model training via frequent learning-updates exchange between edge devices and server. Due to the constrained bandwidth, only a subset of devices can upload their updates at each communication round. This has led to an active research area in FEEL studying the optimal device scheduling policy for minimizing communication time. However, owing to the difficulty in quantifying the exact communication time, prior work in this area can only tackle the problem partially by considering either the communicatio…
