OpenUH: an optimizing, portable OpenMP compiler

Liao, Chunhua; Hernández, Óscar; Chapman, Barbara; Chen, Wenguang; Zheng, Weimin

doi:10.1002/cpe.1174

Cited by 61 publications

(33 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Nested parallelism is specified as an option in the OpenMP and is supported in some implementations such as [22] [37] on nested parallelism for CPUs, our solution is the first language extension to support nested parallelism for GPGPU platforms, and we show different ways to handle the scan and reduction operations, the interwarp and intra-warp NP schemes, as well as careful resource managements on GPGPUs.…”

Section: Related Work and Discussionmentioning

confidence: 99%

CUDA-NP: Realizing Nested Thread-Level Parallelism in GPGPU Applications

Yang

Zhou

2015

J. Comput. Sci. Technol.

View full text Add to dashboard Cite

Parallel programs consist of series of code sections with different thread-level parallelism (TLP). As a result, it is rather common that a thread in a parallel program, such as a GPU kernel in CUDA programs, still contains both sequential code and parallel loops. In order to leverage such parallel loops, the latest Nvidia Kepler architecture introduces dynamic parallelism, which allows a GPU thread to start another GPU kernel, thereby reducing the overhead of launching kernels from a CPU. However, with dynamic parallelism, a parent thread can only communicate with its child threads through global memory and the overhead of launching GPU kernels is non-trivial even within GPUs.In this paper, we first study a set of GPGPU benchmarks that contain parallel loops, and highlight that these benchmarks do not have a very high loop count or high degrees of TLP. Consequently, the benefits of leveraging such parallel loops using dynamic parallelism are too limited to offset its overhead. We then present our proposed solution to exploit nested parallelism in CUDA, referred to as CUDA-NP. With CUDA-NP, we initially enable a high number of threads when a GPU program starts, and use control flow to activate different numbers of threads for different code sections. We implemented our proposed CUDA-NP framework using a directive-based compiler approach. For a GPU kernel, an application developer only needs to add OpenMP-like pragmas for parallelizable code sections. Then, our CUDA-NP compiler automatically generates the optimized GPU kernels. It supports both the reduction and the scan primitives, explores different ways to distribute parallel loop iterations into threads, and efficiently manages on-chip resource. Our experiments show that for a set of GPGPU benchmarks, which have already been optimized and contain nested parallelism, our proposed CUDA-NP framework further improves the performance by up to 6.69 times and 2.18 times on average.

show abstract

Section: Related Work and Discussionmentioning

confidence: 99%

CUDA-NP: Realizing Nested Thread-Level Parallelism in GPGPU Applications

Yang

Zhou

2015

J. Comput. Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…We do it in [7] in an algorithmic fashion, using rewriting rules, that are similar to a LEX-YACC program. The main idea is to use outlining, a widely used technique in the OpenMP community, see [14]. To have a taste of that transformation, one should observe how the MetaFork code of Figure 12 is transformed into the CilkPlus code of Figure 13.…”

Section: Semantics Of the Parallel Constructs In Metaforkmentioning

confidence: 99%

MetaFork: A Framework for Concurrency Platforms Targeting Multicores

Chen

Maza

Shekar

et al. 2014

Using and Improving OpenMP for Devices, Tasks, and More

View full text Add to dashboard Cite

Abstract. We present MetaFork, a metalanguage for multithreaded algorithms based on the fork-join concurrency model and targeting multicore architectures. MetaFork is implemented as a source-to-source compilation framework allowing automatic translation of programs from one concurrency platform to another. The current version of this framework supports CilkPlus and OpenMP. We evaluate the benefits of the MetaFork framework through a series of experiments, such as narrowing performance bottlenecks in multithreaded programs. Our experiments show also that, if a native program, written either in CilkPlus or OpenMP, has little parallelism overhead, then the same property holds for its OpenMP or CilkPlus counterpart translated by MetaFork.

show abstract

“…Representative examples include Omni [8], OdinMP [12] and OpenUH [6]. Most research compilers adopt the source-to-source translation approach.…”

Section: Related Workmentioning

confidence: 99%

A ROSE-Based OpenMP 3.0 Research Compiler Supporting Multiple Runtime Libraries

Liao

Quinlan

Panas

et al. 2010

Beyond Loop Level Parallelism in OpenMP: Accelerators, Tasking and More

Self Cite

View full text Add to dashboard Cite

Abstract. OpenMP is a popular and evolving programming model for shared-memory platforms. It relies on compilers for optimal performance and to target modern hardware architectures. A variety of extensible and robust research compilers are key to OpenMP's sustainable success in the future. In this paper, we present our efforts to build an OpenMP 3.0 research compiler for C, C++, and Fortran; using the ROSE source-tosource compiler framework. Our goal is to support OpenMP research for ourselves and others. We have extended ROSE's internal representation to handle all of the OpenMP 3.0 constructs and facilitate their manipulation. Since OpenMP research is often complicated by the tight coupling of the compiler translations and the runtime system, we present a set of rules to define a common OpenMP runtime library (XOMP) on top of multiple runtime libraries. These rules additionally define how to build a set of translations targeting XOMP. Our work demonstrates how to reuse OpenMP translations across different runtime libraries. This work simplifies OpenMP research by decoupling the problematic dependence between the compiler translations and the runtime libraries. We present an evaluation of our work by demonstrating an analysis tool for OpenMP correctness. We also show how XOMP can be defined using both GOMP and Omni and present comparative performance results against other OpenMP compilers.

show abstract

OpenUH: an optimizing, portable OpenMP compiler

Cited by 61 publications

References 6 publications

CUDA-NP: Realizing Nested Thread-Level Parallelism in GPGPU Applications

CUDA-NP: Realizing Nested Thread-Level Parallelism in GPGPU Applications

MetaFork: A Framework for Concurrency Platforms Targeting Multicores

A ROSE-Based OpenMP 3.0 Research Compiler Supporting Multiple Runtime Libraries

Contact Info

Product

Resources

About