Visualization of Parallel Programming

Visualization of Parallel Programming

A Tool for Understanding Message Passing in Parallel Systems

Andrew Schwartz ‘13Computer ScienceUnion College Advisor: Professor Barr

IntroductionIntel® Quad Core Processor

“Supercomputer”

1

Sum (Single Processor)• Sum first 100

integers

Input1-100

Processor

x=5050

Output

x=1x=x+2x=x+3…x=x+100

100 additions

1

Sum (5 Processors)

Input1-100

xtotal =5050

Output

2

3

4

5

1-20

21-4041-6061-8081-100

x1= 210x2= 610

x4= 1410

x3= 1010

x5= 1810

xtotal = x1+x2xtotal = xtotal +x3 xtotal = xtotal +x4 xtotal = xtotal +x5

100+ 4 additionsFirst step parallelized

Timeline Analysis

0 100

Single processor

Time

(100 additions = 100 time steps)

P0

0 100

Five processors

Time (104 additions, but ~24

time steps

P0

20 24

P5

Architecture Diagrams

Courtesy: Florida State University

NUMA Architecture

goparallel.sourceforge.ne

Topologies and Principles

Background

• Options limited

• Needs more information!

NEXT!

Background• “Installation a

nightmare”• Dozens of views and

diagrams

OVERKILL

MPICL/ParaGraph Overview

???

Implementation

Implementation

Data Analysis

Side by Side

Demo

Future Work

• Core design challenges met• QoL functionality still left to be implemented before release• Extensible project design -> new functions can be modeled

with new algos• Port to UNIX via Mono• Initial results promising but need to test in the classroom