Arenaz slides-booth-talks-sc16-openmp
2 likes818 views
The document discusses the development of Parallware, a software for guided parallelization using OpenMP and OpenACC to simplify parallel programming, which is often a manual and challenging process. It outlines the collaboration between Oak Ridge National Laboratory (ORNL) and Appentra, focusing on tools and training for high-performance computing (HPC). The conclusions emphasize the introduction of the Parallware Trainer at the SC16 showcase and invite feedback through an early access program.
Arenaz slides-booth-talks-sc16-openmp
- 1. LLVM-based Software for Guided Parallelization with OpenMP Manuel Arenaz manuel.arenaz@apppentra.com
- 2. OUTLINE ● Why Parallware? ● ORNL & Appentra ● Parallware Trainer (DEMO) ● Conclusions
- 3. Software modernization through parallelization with MPI+X High-level programming: X = OpenMP or OpenACC WHY DEVELOPING PARALLWARE? PARALLEL PROGRAMMING IS HARD! Currently a manual process Can we make it easier? HPC WORKFLOW
- 4. OUTLINE ● Why Parallware? ● ORNL & Appentra ● Parallware Trainer (DEMO) ● Conclusions
- 5. COLLABORATION ON TOOLS INNOVATION ORNL & APPENTRA ORNL Industrial Partnership Program, project CSC193: “Porting Parallware Tools to Large HPC Installations including Titan” (2015-2016) Fernanda Foertter Oscar Hernandez Jacobo Lobeiras, Manuel Arenaz, Oscar Hernández: Experiences in extending parallware to support OpenACC. WACCPD@SC 2015: 4:1-4:12
- 6. ORNL & APPENTRA This research used resources of the Oak Ridge Leadership Computing Facility, which is a DOE Office of Science User Facility supported under Contract DE-AC05-00OR22725. Validation Deployment Real environment TECHNOLOGY }{ 2x - 40x Titan Microbenchmarks
- 7. ORNL & APPENTRA Newcomers Learning Microbenchmarks TRAINING { } Work on specification of new tools for productive HPC training
- 8. OUTLINE ● Why Parallware? ● ORNL & Appentra ● Parallware Trainer (DEMO) ● Conclusions
- 9. Interactive Tool for HPC Training “Tell me, I will forget, Show me, I may remember, Involve me, I will understand.” - Confucius Emerging Technologies - Higher productivity in HPC training - “Learn by Doing” & “Student-Centric” EXPERIENTIAL LEARNING
- 10. Interactive Tool for HPC Training “Tell me, I will forget, Show me, I may remember, Involve me, I will understand.” - Confucius - Mandelbrot - Sparse Matrix - Vector product - Laplace computations - PGI compiler OpenMP & OpenACC MICROBENCHMARKS
- 11. - Interactive real-time editor GUI - Assisted code parallelization using OpenMP & OpenACC - Programming language C - Detailed report of the parallelism discovered in the code - Support for multiple compilers Interactive Tool for HPC Training “Tell me, I will forget, Show me, I may remember, Involve me, I will understand.” - Confucius “Take-away your work” (based on make & ssh) TECHNICAL FEATURES
- 12. OUTLINE ● Why Parallware? ● ORNL & Appentra ● Parallware Trainer (DEMO) ● Conclusions
- 13. CONCLUSIONS ● Present Parallware Trainer at SC16 Emerging Technologies Showcase ○ Tuesday, Wednesday, Thursday 9:00 AM - 17:30 PM, Room 155-B ○ Also booth talks at DoE, OpenMP and OpenACC ● Parallware Trainer Early Access Program: ○ Invitation to participate (we need feedback) ○ Contact me <manuel.arenaz@appentra.com> ● Continue ORNL & Appentra collaboration: ○ Prototype of Parallware Assistant for HPC developers ○ Search other innovative tools based on Parallware
- 14. CONCLUSIONS ● Present Parallware Trainer at SC16 Emerging Technologies Showcase ○ Tuesday, Wednesday, Thursday 9:00 AM - 17:30 PM, Room 155-B ○ Also booth talks at DoE, OpenMP and OpenACC ● Parallware Trainer Early Access Program: ○ Invitation to participate (we need feedback from users) ○ Contact me <manuel.arenaz@appentra.com> ● Continue ORNL & Appentra collaboration: ○ Prototype of Parallware Assistant for HPC developers ○ Search other innovative tools based on Parallware
- 15. CONCLUSIONS ● Present Parallware Trainer at SC16 Emerging Technologies Showcase ○ Tuesday, Wednesday, Thursday 9:00 AM - 17:30 PM, Room 155-B ○ Also booth talks at DoE, OpenMP and OpenACC ● Parallware Trainer Early Access Program: ○ Invitation to participate (we need feedback) ○ Contact me <manuel.arenaz@appentra.com> ● Continue ORNL & Appentra collaboration: ○ Prototype of Parallware Assistant for HPC developers ○ Search other innovative tools based on Parallware
- 16. LLVM-based Software for Guided Parallelization with OpenMP Manuel Arenaz manuel.arenaz@apppentra.com