Ligra: A Lightweight Graph Processing Framework for Shared Memory J. - - PowerPoint PPT Presentation

ligra a lightweight graph processing framework for shared
SMART_READER_LITE
LIVE PREVIEW

Ligra: A Lightweight Graph Processing Framework for Shared Memory J. - - PowerPoint PPT Presentation

Ligra: A Lightweight Graph Processing Framework for Shared Memory J. Shun and G. Blelloch Presented by Dmitry Kazhdan Overview Background (briefly) Existing work Key contributions Results Criticism Conclusions


slide-1
SLIDE 1
  • J. Shun and G. Blelloch

Ligra: A Lightweight Graph Processing Framework for Shared Memory

Presented by Dmitry Kazhdan

slide-2
SLIDE 2

Overview

  • Background (briefly)
  • Existing work
  • Key contributions
  • Results
  • Criticism
  • Conclusions
  • Questions
slide-3
SLIDE 3

Background

slide-4
SLIDE 4

Background

Motivation:

  • Efficient Graph Processing

Opportunities:

  • Parallelism
  • Large number of cores + RAM in a single server
slide-5
SLIDE 5

Related Work

  • Pregel
  • PowerGraph
  • Green-Marl
  • X-Stream
  • Pegasus
  • GraphLab
slide-6
SLIDE 6

Ligra

slide-7
SLIDE 7

Ligra

  • Graph processing framework
  • Relying on multicore machines with shared memory
  • Offering parallel processing
slide-8
SLIDE 8

Contributions

  • Implemented in memory (single machine)
  • Lightweight (contains a few thousand lines of C++)
  • Easily extendable/customisable
  • Minimal (offering a small number of primitives/abstractions)
  • Ligra+ offers graph compression
slide-9
SLIDE 9

Abstractions

  • Graph and VertexSubset datatypes.
  • EdgeMap and VertexMap functions.
slide-10
SLIDE 10

Use Cases

  • Breadth-first Search
  • Betweenness Centrality
  • Graph Radii Estimation
  • Connected Components
  • Page Rank
  • Bellman-Ford Shortest Paths
slide-11
SLIDE 11

Evaluation

slide-12
SLIDE 12

Evaluation

  • Evaluated algorithm performance on set
  • f selected graphs
  • Showed effects of using more threads
  • Gave performance comparisons to other

systems (albeit brief, often using different setups)

slide-13
SLIDE 13

Evaluation

  • Showed system scalability on randomly generated graphs
slide-14
SLIDE 14

Resources

http://jshun.github.io/ligra/index.html https://github.com/jshun/ligra

slide-15
SLIDE 15

Criticism

slide-16
SLIDE 16

Criticism/Discussion

“and our code is much simpler than theirs”

slide-17
SLIDE 17

Criticism/Discussion

  • How representative are the examples?
  • Too much time spent describing algorithms
slide-18
SLIDE 18

Criticism/Discussion

  • Single experimental setup (“but the results are slower than the ones from the

Intel machine so we only report the latter”)

  • Performance comparisons not detailed or not meaningful (“we achieved faster

results”)

slide-19
SLIDE 19

Conclusions

slide-20
SLIDE 20

Conclusions

  • Introduced Ligra
  • Compared to existing systems
  • Presented evaluation results
  • Criticism
slide-21
SLIDE 21

Questions?