Auto-Tuning of 3D Acoustic Wave Propagation in Shared Memory Environments

Barros, Tiago; Fernandes, J. B.; Souza-de-Assis, I. A.; Xavier‐de‐Souza, Samuel

doi:10.3997/2214-4609.201803072

Cited by 4 publications

(10 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…4 shows that using the OpenMP dynamic scheduler without specifying the chunk size leads to a significantly higher number of cache misses when compared to the other schedulers. This result, as well as the experiments presented by [23], [29], [30], shows that using the OpenMP dynamic scheduler with the default unitary chunk size and other very small chunk sizes for the RTM leads to loss of performance. The reasons for that are the high number of cache misses due to false sharing, and the overhead to manage the distribution of tasks.…”

Section: ) Cache Misses Analysissupporting

confidence: 60%

“…For each CSA iteration, each optimizer only measures the execution time of the first time step in the forward propagation, using its current chunk size (Lines 6 and 13). As shown in [23], the run time of the first time step can accurately represent the total propagation execution time. This first-time step is performed twice (Line 4) and only the elapsed time of the second repetition is registered (Lines 5 and 12) in order to avoid cache population effects.…”

Section: Csa-based Auto-tuningmentioning

confidence: 99%

“…Differently, we employ CSA to automatically determine the size of the chunks of parallel loop iterations. The relation between chunk size and total execution time in geophysical methods is unknown but has been shown to have many local optima [23], [30]. By using a global optimization strategy, our method aims to increase the probability of finding the global optimum.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Auto-Tuning of Dynamic Scheduling Applied to 3D Reverse Time Migration on Multicore Systems

et al. 2020

View full text Add to dashboard Cite

Reverse time migration (RTM) is an algorithm widely used in the oil and gas industry to process seismic data. It is a computationally intensive task that suits well in parallel computers. Methods such as RTM can be parallelized in shared memory systems through scheduling iterations of parallel loops to threads. However, several aspects, such as memory size and hierarchy, number of cores, and input size, make optimal scheduling very challenging. In this paper, we introduce a run-time strategy to automatically tune the dynamic scheduling of parallel loops iterations in iterative applications, such as the RTM, in multicore systems. The proposed method aims to reduce the execution time of such applications. To find the optimal granularity, we propose a coupled simulated annealing (CSA) based auto-tuning strategy that adjusts the chunk size of work that OpenMP parallel loops assign dynamically to worker threads during the initialization of a 3D RTM application. Experiments performed with different computational systems and input sizes show that the proposed method is consistently better than the default OpenMP schedulers, static, auto, and guided, causing the application to be up to 33% faster. We show that the possible reason for this performance is the reduction of cache misses, mainly level L3, and low overhead, inferior to 2%. Having shown to be robust and scalable for the 3D RTM, the proposed method could also improve the performance of similar wave-based algorithms, such as full-waveform inversion (FWI) and other iterative applications.

show abstract

Section: ) Cache Misses Analysissupporting

confidence: 60%

Section: Csa-based Auto-tuningmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Auto-Tuning of Dynamic Scheduling Applied to 3D Reverse Time Migration on Multicore Systems

et al. 2020

View full text Add to dashboard Cite

show abstract

“…asynchronous one-sided WS with distributed decen-work-commu-commu-PGAS MPI global load RTM memory tralized stealing nication nication RMA information Barros et al [24] Andreolli et al [26] Andreolli et al [27] Sena et al [28] x Hofmeyr et al [29] x Tchiboukdjian et al [30] x Imam and Sarkar [31] x x x Khaitan et al [32] x Tesser et al [33] x Tesser et al [34] x Tesser et al [35] x Padoin et al [36] x Padoin et al [37] x Sharma and Kanungo [38] x x x Zheng et al [39] x x x Martinez et al [40] x x x x Khaitan and Mccalley [41] x x x x Mor and Maillard [42] x x x x Li et al [43] x x x x x x Kumar et al [44] x x x x x x Dinan et al [21] x x x x x x Vishnu and Agarwal [49] x…”

Section: Discussionmentioning

confidence: 99%

“…Several authors have proposed strategies to address the load imbalance for shared memory systems. Barros et al [24] introduced a runtime method based on coupled simulated annealing (CSA) [25] to auto-tune the workload ,478 11 15 8 88 12 50 51 518 13 30 24 248 14 62 25 258 15 32 55 558 16 63 57 578 17 17 59 598 18 20 21 387 19 2 1 19 20 23 4 49 21 22 5 59 22 42 8 89 23 28 29 178 24 34 9 478 25 35 12 -26 35 13 139 27 30 14 149 28 21 16 307 29 41 18 318 30 49 37 379 31 13 38 389 32 15 39 399 33 50 54 549 34 18 57 579 35 5 58 589 36 0 59 599 37 3 60 609 38 38 61 619 Table 2: Steal attempts varyi...…”

Section: Related Workmentioning

confidence: 99%

Distributed-Memory Load Balancing With Cyclic Token-Based Work-Stealing Applied to Reverse Time Migration

et al. 2019

View full text Add to dashboard Cite

Reverse time migration (RTM) is a prominent technique in seismic imaging. Its resulting subsurface images are used in the industry to investigate with higher confidence the existence and the conditions of oil and gas reservoirs. Because of its high computational cost, RTM must make use of parallel computers. Balancing the workload distribution of an RTM is a growing challenge in distributed computing systems. The competition for shared resources and the differently-sized tasks of the RTM are some of the possible sources of load imbalance. Although many load balancing techniques exist, scaling up for large problems and large systems remains a challenge because synchronization overhead also scales. This paper proposes a cyclic token-based work-stealing (CTWS) algorithm for distributed memory systems applied to RTM. The novel cyclic token approach reduces the number of failed steals, avoids communication overhead, and simplifies the victim selection and the termination strategy. The proposed method is implemented as a C library using the one-sided communication feature of the message passing interface (MPI) standard. Results obtained by applying the proposed technique to balance the workload of a 3D RTM system present a factor of 14.1 % speedup and reductions of the load imbalance of 78.4 % when compared to the conventional static distribution. 1The migration of seismic data is the process that attempts to build an image of the Earth's interior from recorded field data. Migration places these data into their actual geological position in the subsurface using numerical approximations of either wave-theoretical or ray-theoretical approaches to simulate the propagation of seismic waves [1].The wave-theoretical approach to the propagation of seismic waves employs the finite difference method (FDM) [2,3] to numerically solve the equation describing the movement of the waves [1,4]. This approach is prevalent among the geophysical community, due to its capacity of dealing with substantial velocity variations in complex geology (e.g., pre-salt).Reverse time migration (RTM) [5-9] implements this approach. It is one of the most known FDM-based migration methods. RTM is computationally intensive in terms of data storage and handling, and its use of high-complexity algorithms. Therefore, exploiting parallelism is mandatory for RTM implementations in 3D Earth models (3D RTM) [10].Parallel architectures can be classified as shared memory, when there is a single memory address space available to all processing units (e.g., nodes or cores), or distributed memory otherwise [11]. Many scientific and industrial computational resources are distributed memory systems composed of multiprocessor nodes with shared memory systems. A hybrid parallel application works at these two levels of parallelism. It can distribute the total workload among the nodes of a distributed memory system. Each node, then, distributes its subset of the workload among the processing units of its shared memory system. Parallel machines can also be described as het...

show abstract

Towards a more robust input for stereotomography

Barros

Lopes

Chauris

2022

Geophysical Prospecting

View full text Add to dashboard Cite

Stereotomography is a slope‐based tomographic method that uses seismic reflection data, consisting of traveltimes and slopes, to estimate the subsurface velocity macro‐model. Here, we interpret the macro‐model as the large‐scale components of the acoustic velocity model of the subsurface. Tomographic methods are often solved with local optimization techniques, which are dependent on the initial input data. In that sense, selecting the correct input data by estimating and picking slopes is a crucial part of the stereotomography problem. In this paper, we address the problem of slope estimation and picking in stereotomography. Our central proposal uses the attributes obtained from the common‐offset common‐reflection‐surface method as input data for stereotomography. We use the global optimization method known as differential evolution to estimate these attributes, resulting in gathers of attributes estimated for the complete seismic data. This strategy presents reliable estimates when the seismic data are highly corrupted by random noise or present geological features that may lead to the failure of classical methods. We also propose an automatic picking strategy to extract the traveltimes and slopes from the common‐offset common‐reflection‐surface attributes gathers. We illustrate with synthetic examples the benefits of using the proposed framework for obtaining input data for stereotomography.

show abstract

Auto-Tuning of 3D Acoustic Wave Propagation in Shared Memory Environments

Cited by 4 publications

References 0 publications

Auto-Tuning of Dynamic Scheduling Applied to 3D Reverse Time Migration on Multicore Systems

Auto-Tuning of Dynamic Scheduling Applied to 3D Reverse Time Migration on Multicore Systems

Distributed-Memory Load Balancing With Cyclic Token-Based Work-Stealing Applied to Reverse Time Migration

Towards a more robust input for stereotomography

Contact Info

Product

Resources

About