1 / 21

Fire Benchmark Parallelisation

Fire Benchmark Parallelisation. Programming of Supercomputers WS 11/12 Sam Maurus. What is Fire Benchmark?. CFD solver for arbitrary geometries This project concerned itself with the gccg solver. How Fast is Fire Benchmark Sequentially?. What effect does the input file-format have?.

ross
Télécharger la présentation

Fire Benchmark Parallelisation

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Fire Benchmark Parallelisation Programming of Supercomputers WS 11/12 Sam Maurus

  2. What is Fire Benchmark? • CFD solver for arbitrary geometries • This project concerned itself with the gccg solver

  3. How Fast is Fire Benchmark Sequentially?

  4. What effect does the input file-format have?

  5. Data structures in gccg Points Elements

  6. Data structures in gccg z y ( x points array

  7. Data structures in gccg elems array

  8. Data structures in gccg lcc array

  9. Data distribution approach Process 0 (root) Process 1 • Root Process Tasks: • Read input file • Partition elements using chosen approach • Create and send relevant mapping arrays to each processes • Broadcast common data package to each processor Process 2 Process 3 = lcc, ne, epart, countPart, bs_local, be_local…

  10. Communication model

  11. Communication model P3 P3 has_ghost_neighbour = 0 has_ghost_neighbour = 1 P3 P3 P3 P5 P3 P3 has_ghost_neighbour array

  12. Communication model Process 0 Process 1 Process k (k = count) Process x • Computational loop, phase one: • Start Isend to required processes (where cellCountsToSend[i] > 0) • Start Irecv from required processes (where cellCountsToRecv[i] > 0) • Process local elements that have no ghost neighbours • Wait on all requests • Update remaining local elements

  13. Communication model

  14. Problems overcome MPI_WAIT FUNCTION • Problem:MPI_Wait was being executed both for the send and receive requests for every element processed • Solution:has_ghost_neighbour array introduced, allowing for intermediate computation. MPI_Wait then only called once for each request. BEFORE AFTER

  15. Problems overcome REDUNDANT REPROCESSING OF INPUT FILE • Problem: Input file was being read once at initialisation and again for writing the result (redundant) • Solution: ‘Write solution’ code was refactored to re-use the relevant file information obtained from the first read BEFORE AFTER

  16. Speedup – cojack

  17. Speedup – pent

  18. Speedup – drall

  19. Speedup – tjunc

  20. Speedup – full execution

  21. Thanks for listening  • Discussion time!

More Related