際際滷

際際滷Share a Scribd company logo

OpenACC and Hackathons Monthly Highlights

Download as PPTX, PDF
OpenACC
OpenACC

Stay up-to-date on the latest news, research and resources. This month's edition covers the Georgia Tech Open Hackathon, milestones in OpenACC development, upcoming Open Hackathons and Bootcamps, NVIDIA's developer program, and more!

1 of 13
Download to read offline
February 2023 OPENACC AND OPEN HACKATHONS MONTHLY HIGHLIGHTS
2 WHAT IS OPENACC? main() { <serial code> #pragma acc kernels { <parallel code> } } Add Simple Compiler Directive POWERFUL & PORTABLE Directives-based programming model for parallel computing Designed for performance and portability on CPUs and GPUs SIMPLE Open Specification Developed by OpenACC Organization
3 silica IFPEN, RMM-DIIS on P100 OPENACC SPECIFICATION MOMENTUM Wide Adoption Across Key HPC Codes ANSYS Fluent Gaussian VASP LSDalton MPAS GAMERA GTC XGC ACME FLASH COSMO Numeca 400+ APPS* USING OPENACC Prof. Georg Kresse Computational Materials Physics University of Vienna For VASP, OpenACC is the way forward for GPU acceleration. Performance is similar to CUDA, and OpenACC dramatically decreases GPU development and maintenance efforts. Were excited to collaborate with NVIDIA and PGI as an early adopter of Unified Memory. VASP Top Quantum Chemistry and Material Science Code * Applications in production and development
4 VIEW TIMELINE Take a ride down memory lane and look at the key milestones in the development of the OpenACC specification and supporting tools over the years. OPENACC DEVELOPMENT TIMELINE View the Community-Driven Progress OpenACC made it practical to develop for GPU-based hardware while retaining a single source for almost all the COSMO physics code. Dr. Oliver Fuhrer Senior Scientist MeteoSwiss
5 UPCOMING OPEN HACKATHONS & BOOTCAMPS COMPLETE LIST OF EVENTS Event Call Closes Event Date KISTI AI for Science Bootcamp April 3, 2023 April 11-12, 2023 NCI N-Ways to GPU Programming Bootcamp April 4, 2023 April 13-14, 2023 EuroCC AI for Science Bootcamp March 28, 2023 April 17-18, 2023 NCI End-to-End AI for Science Bootcamp April 10, 2023 April 19-21, 2023 NVIDIA/HLRS SciML GPU Bootcamp April 5, 2023 April 26-27, 2023 MPCDF AI for Science April 25, 2023 May 11-12, 2023 EuroCC N-Ways to GPU Programming Bootcamp April 24, 2023 May 15-16, 2023 CINECA Open Hackathon March 26, 2023 June 12, 19-21, 2023 Princeton Open Hackathon April 5, 2023 June 14, 21-23, 2023 Digital in 2023: Our virtual events continue to offer the same high-touch training and mentorship without the hassle of travel!
6 READ ARTICLE Georgia Tech hosted a virtual open hackathon for participants across academia, national laboratories, and industry to advance and test ongoing software on powerful HPC hardware. On the final day, teams presented research showcasing insightful applications in physics, medicine, computer science, and more. GEORGIA TECH HOSTS HACKATHON Interdisciplinary Teams Optimize Software This hackathon was absolutely worth it. It was a great learning experience and the potential application that we developed for our program is enormous. Ruijia Wang Ph.D. Student, School of Civil and Environmental Engineering Georgia Institute of Technology
7 LEARN MORE For those building new innovations and pushing the envelope of research using AI and HPC, having the right set of tools, training, and support makes all the difference in the quest for success. As a partner of the global Open Hackathons program, NVIDIA can help. Learn how NVIDIA's Developer Program supports the research and developer community by providing access to the latest tools and technology, offering training on their most recent products, and sharing expert help when its needed. NVIDIA REGISTERED DEVELOPER PROGRAM Tools, Training, and Expertise
8 LISTEN NOW Listen to Satoshi Matsuoka from RIKEN Center for Computational Science (R-CCS) and Torsten Hoefler from ETH Zurich offer their thoughts on 12 important topics as outlined in their future-looking paper entitled Myths and Legends in High-Performance Computing that challenges common lines of thought they say have become, to varying degrees, accepted wisdom in HPC, helping to formulate questions and initiate important discussions. ON MYTHS AND LEGENDS OF HPC @HPCPodcast Interviews S. Matsuoka and T. Hoefler
9 RESOURCES According to estimates, between 10 and 25% of the grain crop is lost in the post-harvest. The correct drying of the beans is one of the actions to contain this loss. Drying grains is one of the most critical steps in grain processing for its proper conservation after harvest. As the grain mass is a set of solid and empty spaces (holes) through which a fluid can pass, its drying could be considered a problem of the coupled open-porous medium. In this paper, we propose a mathematical and computer simulation model which describes the convection in a free flow with a porous obstacle applied to the grain drying processing. A computational fluid dynamics scheme was implemented in FORTRAN using Finite Volume to simulate and compute the numerical solutions. The code is parallel implemented using OpenMP (loop and teams approach) and OpenACC programming interfaces. We use three case studies with different mesh sizes to evaluate the implementations. As a result, there was a reduction in processing time in the cases. The total simulation time for a multicore architecture (16 physical cores) was 6,14 times less using parallel loops and 8.5 using parallel teams, and 8.38 using a single GPU (Quadro M5000) READ PAPER Paper: Parallel OpenMP and OpenACC Porous Media Simulation H鱈gor U辿linton da Silva, Natiele Lucca, Claudio Schepke, Dalmo Paim de Oliveira, and C辿sar Flaubiano da Cruz Cristaldo Figure 6
10 RESOURCES OpenACC is designed to offer performance portability across CPUs with SIMD extensions and accelerators based on GPU or manycore architecture. We are working on the design of OpenACC compiler for A64FX manycore processor with Arm SVE. We use a source-to- source compiler infrastructure, Omni compiler, to generate the code with OpenMP SIMD directive and generate a translated code in OpenCL which is executed by OpenCL runtime for CPUs. We expect OpenACC to be used to exploit different levels of parallelism for NUMA node, cores in a NUMA node and SIMD vector by Arm SVE in A64FX manycore processor, more than conventional flat OpenMP parallel loops. The preliminary results of our source-to-source transformation with OpenMP SIMD directive is presented. READ PAPER Paper: OpenACC Execution Models for Manycore Processor with ARM SVE Mitsuhisa Sato and Miwako Tsuji Figure 1. Block diagram of A64FX manycore processor.
11 RESOURCES Website: OpenHackathons.org Technical Resources VISIT SITE Explore a wealth of resources for parallelization and accelerated computing across HPC, AI and Big Data. Review a collection of videos, presentations, GitHub repos, tutorials, libraries, and more to help you advance your skills and expand your knowledge.
12 STAY IN THE KNOW: JOIN THE COMMUNITY OPENACC AND HACKATHON UPDATES JOIN TODAY The OpenACC Organization is dedicated to helping the research and developer community advance science by expanding their accelerated and parallel computing skills. Take an active role in influencing the future of both the OpenACC specification and the organization itself by becoming a member of the community. Keep abreast of the new tools, latest resources, recent research, and upcoming events.
WWW.OPENACC.ORG Learn more at

More Related Content

OpenACC and Hackathons Monthly Highlights

  • 1. February 2023 OPENACC AND OPEN HACKATHONS MONTHLY HIGHLIGHTS
  • 2. 2 WHAT IS OPENACC? main() { <serial code> #pragma acc kernels { <parallel code> } } Add Simple Compiler Directive POWERFUL & PORTABLE Directives-based programming model for parallel computing Designed for performance and portability on CPUs and GPUs SIMPLE Open Specification Developed by OpenACC Organization
  • 3. 3 silica IFPEN, RMM-DIIS on P100 OPENACC SPECIFICATION MOMENTUM Wide Adoption Across Key HPC Codes ANSYS Fluent Gaussian VASP LSDalton MPAS GAMERA GTC XGC ACME FLASH COSMO Numeca 400+ APPS* USING OPENACC Prof. Georg Kresse Computational Materials Physics University of Vienna For VASP, OpenACC is the way forward for GPU acceleration. Performance is similar to CUDA, and OpenACC dramatically decreases GPU development and maintenance efforts. Were excited to collaborate with NVIDIA and PGI as an early adopter of Unified Memory. VASP Top Quantum Chemistry and Material Science Code * Applications in production and development
  • 4. 4 VIEW TIMELINE Take a ride down memory lane and look at the key milestones in the development of the OpenACC specification and supporting tools over the years. OPENACC DEVELOPMENT TIMELINE View the Community-Driven Progress OpenACC made it practical to develop for GPU-based hardware while retaining a single source for almost all the COSMO physics code. Dr. Oliver Fuhrer Senior Scientist MeteoSwiss
  • 5. 5 UPCOMING OPEN HACKATHONS & BOOTCAMPS COMPLETE LIST OF EVENTS Event Call Closes Event Date KISTI AI for Science Bootcamp April 3, 2023 April 11-12, 2023 NCI N-Ways to GPU Programming Bootcamp April 4, 2023 April 13-14, 2023 EuroCC AI for Science Bootcamp March 28, 2023 April 17-18, 2023 NCI End-to-End AI for Science Bootcamp April 10, 2023 April 19-21, 2023 NVIDIA/HLRS SciML GPU Bootcamp April 5, 2023 April 26-27, 2023 MPCDF AI for Science April 25, 2023 May 11-12, 2023 EuroCC N-Ways to GPU Programming Bootcamp April 24, 2023 May 15-16, 2023 CINECA Open Hackathon March 26, 2023 June 12, 19-21, 2023 Princeton Open Hackathon April 5, 2023 June 14, 21-23, 2023 Digital in 2023: Our virtual events continue to offer the same high-touch training and mentorship without the hassle of travel!
  • 6. 6 READ ARTICLE Georgia Tech hosted a virtual open hackathon for participants across academia, national laboratories, and industry to advance and test ongoing software on powerful HPC hardware. On the final day, teams presented research showcasing insightful applications in physics, medicine, computer science, and more. GEORGIA TECH HOSTS HACKATHON Interdisciplinary Teams Optimize Software This hackathon was absolutely worth it. It was a great learning experience and the potential application that we developed for our program is enormous. Ruijia Wang Ph.D. Student, School of Civil and Environmental Engineering Georgia Institute of Technology
  • 7. 7 LEARN MORE For those building new innovations and pushing the envelope of research using AI and HPC, having the right set of tools, training, and support makes all the difference in the quest for success. As a partner of the global Open Hackathons program, NVIDIA can help. Learn how NVIDIA's Developer Program supports the research and developer community by providing access to the latest tools and technology, offering training on their most recent products, and sharing expert help when its needed. NVIDIA REGISTERED DEVELOPER PROGRAM Tools, Training, and Expertise
  • 8. 8 LISTEN NOW Listen to Satoshi Matsuoka from RIKEN Center for Computational Science (R-CCS) and Torsten Hoefler from ETH Zurich offer their thoughts on 12 important topics as outlined in their future-looking paper entitled Myths and Legends in High-Performance Computing that challenges common lines of thought they say have become, to varying degrees, accepted wisdom in HPC, helping to formulate questions and initiate important discussions. ON MYTHS AND LEGENDS OF HPC @HPCPodcast Interviews S. Matsuoka and T. Hoefler
  • 9. 9 RESOURCES According to estimates, between 10 and 25% of the grain crop is lost in the post-harvest. The correct drying of the beans is one of the actions to contain this loss. Drying grains is one of the most critical steps in grain processing for its proper conservation after harvest. As the grain mass is a set of solid and empty spaces (holes) through which a fluid can pass, its drying could be considered a problem of the coupled open-porous medium. In this paper, we propose a mathematical and computer simulation model which describes the convection in a free flow with a porous obstacle applied to the grain drying processing. A computational fluid dynamics scheme was implemented in FORTRAN using Finite Volume to simulate and compute the numerical solutions. The code is parallel implemented using OpenMP (loop and teams approach) and OpenACC programming interfaces. We use three case studies with different mesh sizes to evaluate the implementations. As a result, there was a reduction in processing time in the cases. The total simulation time for a multicore architecture (16 physical cores) was 6,14 times less using parallel loops and 8.5 using parallel teams, and 8.38 using a single GPU (Quadro M5000) READ PAPER Paper: Parallel OpenMP and OpenACC Porous Media Simulation H鱈gor U辿linton da Silva, Natiele Lucca, Claudio Schepke, Dalmo Paim de Oliveira, and C辿sar Flaubiano da Cruz Cristaldo Figure 6
  • 10. 10 RESOURCES OpenACC is designed to offer performance portability across CPUs with SIMD extensions and accelerators based on GPU or manycore architecture. We are working on the design of OpenACC compiler for A64FX manycore processor with Arm SVE. We use a source-to- source compiler infrastructure, Omni compiler, to generate the code with OpenMP SIMD directive and generate a translated code in OpenCL which is executed by OpenCL runtime for CPUs. We expect OpenACC to be used to exploit different levels of parallelism for NUMA node, cores in a NUMA node and SIMD vector by Arm SVE in A64FX manycore processor, more than conventional flat OpenMP parallel loops. The preliminary results of our source-to-source transformation with OpenMP SIMD directive is presented. READ PAPER Paper: OpenACC Execution Models for Manycore Processor with ARM SVE Mitsuhisa Sato and Miwako Tsuji Figure 1. Block diagram of A64FX manycore processor.
  • 11. 11 RESOURCES Website: OpenHackathons.org Technical Resources VISIT SITE Explore a wealth of resources for parallelization and accelerated computing across HPC, AI and Big Data. Review a collection of videos, presentations, GitHub repos, tutorials, libraries, and more to help you advance your skills and expand your knowledge.
  • 12. 12 STAY IN THE KNOW: JOIN THE COMMUNITY OPENACC AND HACKATHON UPDATES JOIN TODAY The OpenACC Organization is dedicated to helping the research and developer community advance science by expanding their accelerated and parallel computing skills. Take an active role in influencing the future of both the OpenACC specification and the organization itself by becoming a member of the community. Keep abreast of the new tools, latest resources, recent research, and upcoming events.