An Evaluation of Auto-Scoping in OpenMP

Voss, Michael; Chiu, Eric M.P.; Chow, Patrick Man Yan; Wong, Catherine; Yuen, Kevin C.J.

doi:10.1007/978-3-540-31832-3_9

Cited by 7 publications

(4 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are several research work on auto-scoping of variables for OpenMP (Lin et al 2004;Royuela et al 2012;Voss et al 2005), the difference with the variable privatization analysis we proposed is the memory model oriented. The auto-scoping of variables for OpenMP is based on shared main memory, but our approach is based on both shared main memory and scratchpad memory.…”

Section: Static Analysismentioning

confidence: 99%

An automatic mapping technique for OpenACC kernel code based on deeply fused and heterogeneous many-core architecture

et al. 2020

View full text Add to dashboard Cite

Now the OpenACC has become a popular programming interface for many-core application programming. Internationally, a lot of research have been done on OpenACC for CPU + GPU heterogeneous many-core architecture. Among them, the PGI OpenACC compiler developed by NVIDIA is the most advanced one. But there are few research on OpenACC related to the Home Grown Heterogeneous Many-Core (HGHM) Architecture that is different from GPU. This paper proposes an automatic mapping technique for OpenACC kernel code based on the OpenACC compiler to a heterogeneous and deeply fused many-core architecture. Our approach uses the static analysis and feedback dynamic analysis of the compiler to perform the automatic mapping of the program parallel kernel code to many-core devices, and it greatly improves the transformation quality of the compiler. Experimental results show that this technique can greatly improve the efficiency of using OpenACC to port applications to heterogeneous and fused many-core system without impacting program acceleration performance.

show abstract

Section: Static Analysismentioning

confidence: 99%

An automatic mapping technique for OpenACC kernel code based on deeply fused and heterogeneous many-core architecture

et al. 2020

View full text Add to dashboard Cite

show abstract

“…They showed that automatic and user-defined clauses can obtain the same performance. Voss et al [25] evaluated the impact of the same clause for auto-parallelization purposes, concluding that several regions cannot be automatically parallelized because of the limitations of the technique: it only works for the OpenMP thread model, it offers limited inter-procedural analysis for arrays, and it lacks support for API functions calls. Later, Royuela et al [17] proposed an algorithm to automatically scope variables in OpenMP 3.1 task regions.…”

Section: Related Workmentioning

confidence: 99%

Static Analysis to Enhance Programmability and Performance in OmpSs-2

Munera

Royuela

Ferrer

et al. 2020

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Task-based parallel programming models based on compiler directives have proved their effectiveness at describing parallelism in High-Performance Computing (HPC) applications. Recent studies show that cutting-edge Real-Time applications, such as those for unmanned vehicles, can successfully exploit these models. In this scenario, OpenMP is a de facto standard for HPC, and is being studied for Real-Time systems due to its time-predictability and delimited functional safety. However, changes in OpenMP take time to be standardized because it sweeps along a large community. OmpSs, instead, is a task-based model for fastprototyping that has been a forerunner of OpenMP since its inception. OmpSs-2, its successor, aims at the same goal, and defines several features that can be introduced in future versions of OpenMP. This work targets compiler-based optimizations to enhance the programmability and performance of OmpSs-2. Regarding the former, we present an algorithm to determine the data-sharing attributes of OmpSs-2 tasks. Regarding the latter, we introduce a new algorithm to automatically release OmpSs-2 task dependencies before a task has completed. This work evaluates both algorithms in a set of well-known benchmarks, and discusses their applicability to the current and future specifications of OpenMP.

show abstract

“…They proved that the performance obtained by the two versions is the same. Voss et al [9] presented an evaluation of auto-scoping in OpenMP based on the work of Lin et al In this context, they observed significant limitations in the usability of the auto-scoping method for automatic parallelization techniques. Implementing the same AUTO data-sharing attribute as Lin et al did, but in the Polaris parallelizing compiler, they evaluated their implementation with a subset of the SPEC benchmark suite.…”

Section: Motivation and Related Workmentioning

confidence: 99%

Auto-scoping for OpenMP Tasks

Royuela

Durán

Liao

et al. 2012

OpenMP in a Heterogeneous World

View full text Add to dashboard Cite

Abstract. Different analysis and optimisations have been devised for OpenMP over the years. However, as OpenMP added asynchronous parallelism in the form of tasks several of them need to be revisited. One such analysis is auto-scoping, the process of automatically determine the data-sharing attributes of variables. Auto-scoping relieved the programmer of determining the data-sharing of variables used in worksharing or parallel regions. Based on the previous work for worksharing and parallel regions, we present an auto-scoping algorithm to work with OpenMP tasks. This is a much complex challenge due to the uncertainty of when a task will be executed, which makes harder to determine with what other parts of the program will be executed concurrently. We also present an implementation of the algorithm and results with a number of benchmarks showing that the algorithm is able to correctly scope a large percentage of the variables present in them.

show abstract

An Evaluation of Auto-Scoping in OpenMP

Cited by 7 publications

References 9 publications

An automatic mapping technique for OpenACC kernel code based on deeply fused and heterogeneous many-core architecture

An automatic mapping technique for OpenACC kernel code based on deeply fused and heterogeneous many-core architecture

Static Analysis to Enhance Programmability and Performance in OmpSs-2

Auto-scoping for OpenMP Tasks

Contact Info

Product

Resources

About