OpenCL vs OpenACC: Lessons from Development of Lattice QCD Simulation Code

Matsufuru, Hideo; Aoki, Sinya; Aoyama, T.; Kanaya, K.; Motoki, Shinji; Namekawa, Yusuke; Nemura, Hidekatsu; Taniguchi, Yoshiaki; Ueda, Satoru; Ukita, N.

doi:10.1016/j.procs.2015.05.316

Cited by 13 publications

(5 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is an update of the result reported in Ref. [10,11]. The run-time parameters for thread grouping are adjusted for each device.…”

Section: Performancementioning

confidence: 96%

“…Among available frameworks, OpenCL and OpenACC are attractive because of their portability. Application of OpenCL [7,8,9,10,11] and OpenACC [12,11] to lattice QCD have been started. We note these two frameworks have contrasting features: OpenCL uses explicit API-based controls of the device, while OpenACC is directive-based and a compiler generates procedures to use the device.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Lattice QCD code set Bridge++ on arithmetic accelerators

Motoki¹,

Aoki²,

Aoyama³

et al. 2016

Proceedings of the 33rd International Symposium on Lattice Field Theory — PoS(LATTICE 2015)

View full text Add to dashboard Cite

“…This is an update of the result reported in Ref. [10,11]. The run-time parameters for thread grouping are adjusted for each device.…”

Section: Performancementioning

confidence: 96%

Section: Introductionmentioning

confidence: 99%

Lattice QCD code set Bridge++ on arithmetic accelerators

Motoki¹,

Aoki²,

Aoyama³

et al. 2016

Proceedings of the 33rd International Symposium on Lattice Field Theory — PoS(LATTICE 2015)

View full text Add to dashboard Cite

“…the linear equation solver, while keeping the framework and measurement codes of Bridge++ available. This strategy was previously adopted to use GPUs with OpenCL and Ope-nACC [6,7]. We apply this extended-Bridge++ approach to KNL with modifications proper to the many-core and SIMD architecture.…”

Section: Implementation Detailsmentioning

confidence: 99%

Wilson and Domainwall Kernels on Oakforest-PACS

Kanamori

Matsufuru

2018

EPJ Web Conf.

View full text Add to dashboard Cite

We report the performance of Wilson and Domainwall Kernels on a new Intel Xeon Phi Knights Landing based machine named Oakforest-PACS, which is co-hosted by University of Tokyo and Tsukuba University and is currently fastest in Japan. This machine uses Intel Omni-Path for the internode network. We compare performance with several types of implementation including that makes use of the Grid library. The code is incorporated with the code set Bridge++.

show abstract

“…Recent supercomputers, however, adopt a variety of architecture: multi-core parallel machines with wide SIMD (A64FX and Intel processors), and clusters with accelerator devices such as GPUs, PEZY-SC, and vector processors (NEC SX-Aurora). Soon after the first public release of Bridge++ in 2012 [2], we had started to investigate possible extensions of our code to exploit these new architectures while keeping the readability and portability, as well as to develop tuning techniques for them [3,4,5,6,7,8]. Recently we have constructed a framework to incorporate the tuned codes as an alternative part to the previously developed Bridge++ code, and decided to release it as version 2.0.…”

Section: Introductionmentioning

confidence: 99%

General purpose lattice QCD code set Bridge++ 2.0 for high performance computing

Akahoshi

Aoki

Aoyama

et al. 2022

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

Bridge++ is a general-purpose code set for a numerical simulation of lattice QCD aiming at a readable, extensible, and portable code while keeping practically high performance. The previous version of Bridge++ is implemented in double precision with a fixed data layout. To exploit the high arithmetic capability of new processor architecture, we extend the Bridge++ code so that optimized code is available as a new branch, i.e., an alternative to the original code. This paper explains our strategy of implementation and displays application examples to the following architectures and systems: Intel AVX-512 on Xeon Phi Knights Landing, Arm A64FX-SVE on Fujitsu A64FX (Fugaku), NEC SX-Aurora TSUBASA, and GPU cluster with NVIDIA V100.

show abstract

OpenCL vs OpenACC: Lessons from Development of Lattice QCD Simulation Code

Cited by 13 publications

References 4 publications

Lattice QCD code set Bridge++ on arithmetic accelerators

Lattice QCD code set Bridge++ on arithmetic accelerators

Wilson and Domainwall Kernels on Oakforest-PACS

General purpose lattice QCD code set Bridge++ 2.0 for high performance computing

Contact Info

Product

Resources

About