Spatial Sorting Algorithms for Parallel Computing in Networks Max - - PowerPoint PPT Presentation
Spatial Sorting Algorithms for Parallel Computing in Networks Max OrHai, Christof Teuscher 2011 October 3 1 Overview Bubble sort as a Insertion sort in a particle system random network Hypothesis: Spatial abstractions can help
1
particle system
random network
2
(e.g. Lindgren and Nordahl 1990)
(Kennedy and Eberhart 1995)
(Dorigo 1992)
Lose, and Senot, SASO 2010)
3
particles in a simulated continuous space
conditional collisions
partitioned like CA for parallel processing
time space
4
sorting is a natural extension
be non-deterministic without global synchrony
factors beyond particle count: speed, size of space...
time steps: 5 10
10,000 particles
– redness + + blueness – velocity in particle widths per step: 0.05 time steps to sort entire system number of particles in a single-axis space 40 20 60 100 200 300 0.1 0.3 0.2 0.4
5
(as a developmental dataflow program in an amorphous spatial computer)
(Coore 1999)
(e.g. Bachrach, Beal 2006)
Logic Automata (Gershenfeld et al 2010)
6
spatial computer assumptions and terminology
items to be sorted
7
example sequence: extension
8
example sequence: swelling
9
10
11
amorphous network approximates a 2D manifold
12
node count neighbor count
performance and limitations
O(n) time complexity
failure in this model
less efficient use of space than (e.g.) spanning trees
200 400 600 800 75 150 225 300
time steps to sort number of data elements
13
Thanks to the Maseeh College of Engineering and Computer Science Undergraduate Research and Mentoring Program All software models are available:
14