1 / 12

Cache Friendly Parallel Priority Queue

Cache Friendly Parallel Priority Queue. Dinesh Agarwal Distributed Mobile Systems Lab ( DiMoS ) . Overview. Parallel Priority Queue Task Allocation Strategies Prefetching Experimental Results. Parallel Priority Queue. Parallel Binary Heap r elements per node Parallel Heap Property

tangia
Télécharger la présentation

Cache Friendly Parallel Priority Queue

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. Cache Friendly Parallel Priority Queue Dinesh Agarwal Distributed Mobile Systems Lab (DiMoS)

  2. Overview • Parallel Priority Queue • Task Allocation Strategies • Prefetching • Experimental Results

  3. Parallel Priority Queue • Parallel Binary Heap • r elements per node • Parallel Heap Property • Top priority r items at root • General (worker) Processors • Maintenance Processors

  4. Parallel Priority Queue

  5. Task Allocation Strategies • Round Robin • After updating current level, a processor i jumps to Xthlevel; X =(current level + 2 * processors) • Blocking • After updating current level, a processor jumps to Xth level; X = (current level + 2) • Round Robin with Shift • Processor starts working at a level it worked on during last iteration

  6. Prefetching • Observation: To update a node you need its children nodes • Prediction: If the next node to be operated on is known, so are the children • Touch the memory blocks where the children are stored • Overlap these memory reads with work

  7. Experimental Results • Speedup for varying granularity:

  8. Experimental Results • Performance for varying n

  9. Experimental Results • Speedup for varying r:

  10. Experimental Results • Performance on different Machines:

  11. Experimental Results • Prefetching:

  12. Experimental Results • Allocation Strategies:

More Related