Global Task Data-Dependencies in PGAS Applications

Schuchart, Joseph; Gracia, José

doi:10.1007/978-3-030-20656-7_16

Cited by 7 publications

(7 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We have implemented a proof-of-concept (PoC) of continuations within Open MPI. 5 We have provided micro-benchmarks as part of the initial paper [17] and refer interested readers to it default for details. Since then, we have added support for the info keys described in Section 3.5, which should only have a marginal impact on the overhead of continuation registration and execution.…”

Section: Methodsmentioning

confidence: 99%

“…Applications are expressed in terms of work-packages (tasks) with well-defined inputs and outputs, which guide the runtime scheduler in determining the correct execution order of tasks. A wide range of task-based programming models have been developed, ranging from node-local approaches such as OpenMP [1] and OmpSs [2] to distributed systems such as HPX [3], StarPU [4], DASH [5], and PaRSEC [6]. All of them have in common that a set of tasks is scheduled for execution onto a set of worker threads based on constraints provided by the user, either in the form of dependencies or dataflow expressions.…”

Section: Background and Motivationmentioning

confidence: 99%

See 1 more Smart Citation

Callback-based Completion Notification using MPI Continuations

Schuchart,

Samfass,

Niethammer

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Asynchronous programming models (APM) are gaining more and more traction, allowing applications to expose the available concurrency to a runtime system tasked with coordinating the execution. While MPI has long provided support for multi-threaded communication and non-blocking operations, it falls short of adequately supporting APMs as correctly and efficiently handling MPI communication in different models is still a challenge. We have previously proposed an extension to the MPI standard providing operation completion notifications using callbacks, so-called MPI Continuations. This interface is flexible enough to accommodate a wide range of different APMs.In this paper, we present an extension to the previously described interface that allows for finer control of the behavior of the MPI Continuations interface. We then present some of our first experiences in using the interface in the context of different applications, including the NAS parallel benchmarks, the PaRSEC task-based runtime system, and a load-balancing scheme within an adaptive mesh refinement solver called ExaHyPE. We show that the interface, implemented inside Open MPI, enables lowlatency, high-throughput completion notifications that outperform solutions implemented in the application space.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Background and Motivationmentioning

confidence: 99%

Callback-based Completion Notification using MPI Continuations

Schuchart,

Samfass,

Niethammer

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Schuchart et al [32] present the concept of global dependencies among tasks that run on different processes and based on the DASH PGAS model [33]. Tasks can declare both local and remote dependencies so that interprocess synchronizations can be defined as dependencies among tasks at different processes.…”

Section: Related Workmentioning

confidence: 99%

Combining One-Sided Communications with Task-Based Programming Models

Sala

Macià

Beltrán

2021

2021 IEEE International Conference on Cluster Computing (CLUSTER)

View full text Add to dashboard Cite

Hybrid programming combining task-based and message-passing models is an increasingly popular technique to exploit multi-core clusters. The Task-Aware MPI (TAMPI) library integrates both models enabling the safe overlap of computation and communication tasks using two-sided MPI communications. Two-sided primitives combine data transfers with implicit synchronizations, but one-sided models usually offer more efficient data transfers decoupling synchronizations. MPI offers four distinct one-sided synchronization modes, while GASPI is a PGAS API providing one-sided operations with remote notifications for fine inter-process synchronizations.In this paper, we study the challenges of integrating MPI and GASPI one-sided operations with the OpenMP and OmpSs-2 tasking models. We propose and implement several extensions to the GASPI and OmpSs-2 programming models, which are leveraged by a new library called Task-Aware GASPI (TAGASPI). The TAGASPI library allows the efficient and safe use of one-sided operations with remote notifications inside tasks. Both TAGASPI and TAMPI transparently manage communications issued by tasks and allow these to overlap with computation tasks naturally, following a data-flow model. These libraries are complementary and can be mixed in the same application.Our experience porting several mini-apps to this hybrid model shows that TAGASPI helps leverage one-sided communications with similar complexity to pure and hybrid two-sided MPI approaches. We show that our hybrid one-sided approach outperforms the pure MPI strategies, but it also surpasses the TAMPI's performance when stressing communication phases, e.g., increasing the communication parallelism and reducing the communication tasks' sizes.

show abstract

“…The interface proposed here is especially well-suited for nodelocal asynchronous programming models such as OpenMP [40], OmpSs-2 [9], and Intel TBB [43] or in combination with cooperatively scheduled fiber libraries such as Argobots [51] or Qthreads [61] to ease the communication management burden users are facing. However, it may also be useful in distributed runtime systems such as PaRSEC [7], HPX [27], StarPU [3], and DASH [49] to simplify internal communication handling. An evaluation of the integration with runtime systems remains as future work.…”

Section: Related Workmentioning

confidence: 99%

“…Applications are expressed in terms of work-packages (tasks) with well-defined inputs and outputs, which guide the runtime scheduler in determining the correct execution order of tasks. A wide range of task-based programming models have been developed, ranging from node-local approaches such as OpenMP [40] and OmpSs [13] to distributed systems such as HPX [27], StarPU [3], DASH [49], and PaRSEC [7]. All of them have in common that a set of tasks is scheduled for execution by a set of worker threads based on constraints provided by the user, either in the form of dependencies or dataflow expressions.…”

mentioning

confidence: 99%

Fibers are not (P)Threads

Schuchart¹,

Niethammer²,

Gracia³

2020

27th European MPI Users' Group Meeting

Self Cite

View full text Add to dashboard Cite

Asynchronous programming models (APM) are gaining more and more traction, allowing applications to expose the available concurrency to a runtime system tasked with coordinating the execution. While MPI has long provided support for multi-threaded communication and non-blocking operations, it falls short of adequately supporting APMs as correctly and efficiently handling MPI communication in different models is still a challenge. Meanwhile, new low-level implementations of lightweight , cooperatively scheduled execution contexts (fibers, aka user-level threads (ULT)) are meant to serve as a basis for higher-level APMs and their integration in MPI implementations has been proposed as a replacement for traditional POSIX thread support to alleviate these challenges. In this paper, we first establish a taxonomy in an attempt to clearly distinguish different concepts in the parallel software stack. We argue that the proposed tight integration of fiber implementations with MPI is neither warranted nor beneficial and instead is detrimental to the goal of MPI being a portable communication abstraction. We propose MPI Continuations as an extension to the MPI standard to provide callback-based notifications on completed operations, leading to a clear separation of concerns by providing a loose coupling mechanism between MPI and APMs. We show that this interface is flexible and interacts well with different APMs, namely OpenMP detached tasks, OmpSs-2, and Argobots.

show abstract

Global Task Data-Dependencies in PGAS Applications

Cited by 7 publications

References 25 publications

Callback-based Completion Notification using MPI Continuations

Callback-based Completion Notification using MPI Continuations

Combining One-Sided Communications with Task-Based Programming Models

Fibers are not (P)Threads

Contact Info

Product

Resources

About