Dynamic Load Distributions for Adaptive Computations on MIMD Machines using Hybrid Genetic Algorithms (a subset)

Size: px
Start display at page:

Download "Dynamic Load Distributions for Adaptive Computations on MIMD Machines using Hybrid Genetic Algorithms (a subset)"

Transcription

1 Dynamic Load Distributions for Adaptive Computations on MIMD Machines using Hybrid Genetic Algorithms (a subset) TIMOTHY H. KAISER PhD, Computer Science - UNM MS, Electrical Engineering - (Applied Physics) UCSD BS, Physics -UMR

2 Introduction Difficult talk to put together Subset of full dissertation defense Full dissertation can be found at Discuss parallel adaptive grid programs Dynamic load balancing - requirements for effectiveness Scope of research effort Genetic algorithms Sine qua nons for GA effectiveness Development Application 2

3 What are adaptive grid programs? Simulate a region of space and time using a grid of cells Often solving a PDE Weather modeling, fluid flow, aerodynamics Regions where higher spatial fidelity are required, have finer gridding Would want finer gridding at a weather storm front Would want finer gridding at a shock front New regions requiring higher spatial fidelity can move or be created Tornado Explosion Also called AMR or adaptive mesh refinement programs 3

4 4 An example grid program Numerical solution to Euler s equations U F G t x y + + =0 U u v E F u u p uv Eu pu G v uv v p Ev pv = = + + = + + ρ ρ ρ ρ ρ ρ ρ ρ ρ,, 2 2 ρ=density u=x velocity v=y velocity E= total energy These equations model the flow of fluids and gas

5 Parallelizing grid programs for speed up Assign portions of the grid to various processors Requires: Communication between processors Desirable Good load balanced Low communication cost Problem Getting perfect load balance and optimal communication is NP-Hard Various heuristics are used to solve this problem 5

6 Dynamic load balancing of adaptive grid programs Why is dynamic load balancing important? May start with balanced load and low communication The gird is refined in some region Processor holding refined region becomes heavily loaded Need to dynamically balance to get good performance Why is this difficult? Arbitrarily moving regions to new processors can cause high communication Must be done again when the grid changes Still NP-Hard 6

7 Scope of this research effort Needed to select a subset of all problems on which to work Adaptive grid program solves: Euler s fluid equations 2 dimensions Lower levels of parallelism Up to 16 nodes connected together by high speed network (SP1 and SP2) The committee felt such configurations are readily available Research should be applicable to other problems and architectures For very large problems framework would need modification 7

8 Thesis statement A genetic algorithm can be effectively used to decrease the run time of adaptive grid programs by maintaining good load balance and maintaining good communication performance. 8

9 Tools developed for thesis verification PLIFE - framework for a parallel adaptive grid or mesh program Darwin - framework for a parallel genetic algorithm Combined the two to create a parallel adaptive grid hydrodynamics simulation with dynamic load balancing and communication reduction 9

10 What is a Genetic Algorithm? An suboptimization system Find good, but maybe not optimal, solutions to difficult problems Often used on NP-Hard or combinatorial optimization problems Requirements Solution(s) to the problem represented as a string A fitness function Takes as input the solution string Output the desirability of the solution A method of combining solution strings to generate new solutions 10

11 More details on Genetic Algorithms Find solutions to problems by Darwinian evolution Potential solutions are thought of a living entities in a population The strings are the genetic codes of the individuals Individuals are evaluated for their fitness The fittest individuals are allowed to live and sexually reproduce There may be some mutation Parents die and kids start the next generation 11

12 Use of the GA After cells split, GA does the assignment of cells to processors For a particular potential assignment of cells to processors the GA: Counts the communication cost of the distribution Measures the imbalance of the distribution Fitness = weight of communication* (communication function) + weight of balance*(balance function) GAs have not been used in the past because they are too slow 12

13 For the GA to be effective Run fast Return good solutions Good load balance Low communication Key finding of dissertation: To enable fast and good solutions observe 2 sine qua nons Improvements must be made to the basic GA algorithm Domain specific knowledge must be used 13

14 Simulation used for the development of the sine qua nons Used a four node SP1 calculation Initial grid is 128 x 128 Each processor held a corner of the grid Exploded a bomb in the bottom left corner of the grid Restrictions One level of adaptation Only split cells are moved by the GA Cells are moved in clusters of 16 cells 14

15 First sine qua non - Use of domain specific knowledge Larger contiguous blocks cause less communication Clusters of cells are generated at irregular rates Future load can be estimated The most important piece of domain specific knowledge... 15

16 Cells split in the region of the shock front Shock wave propagating to the right. Boundary between split and unsplit cells across which communication will occur if the split cells are moved. 16

17 Cells split in the region of the shock front Implication Cells behind the shock split and may have been moved by the GA Cells in front of shock are not yet split and not yet moved by GA Conflict To balance load GA wants to move split cells To maintain low communication GA wants to not move cells Note: Next time GA is called more cells will have split Because shock propagates, cells in front of the shock will soon split GA will not want to move cells if those to the left have not moved 17

18 Exploitation of this domain specific knowledge Don t count communication between split and unsplit cells Sort term effect: Increases communication time GA is freer to move cells to balance load Long term effect: Contiguous blocks of cells on the same processor are larger Dramatic decrease in communication time 18

19 Results Without using this domain specific knowledge Communication time = 348 seconds Run time = 1510 seconds With domain specific knowledge in use Communication time = 228 seconds Run time = 1390 seconds Run time reduced by 2 minutes Show Movies 19

20 Second sine qua non - Improve the GA algorithm Run GA as a parallel application Developed a fast mutation methodology Use the Mansour-Fox algorithm Improve the Mansour-Fox algorithm 20

21 The Mansour-Fox algorithm Developed for static allocations Variable fitness function Initially put more weight on communication minimization In the end put more weight on balanced load About 50% - 75% through generations do the following: Call hill climbing routine for each member of GA population Greedy algorithm If cell has neighbors on different processor find best processor Increase mutation rate Mutate sections of gene which represent cells on boundaries 21

22 Mansour-Fox algorithm continued Problems: Slow because the hill climbing routine is slow Can not terminate early because of variable fitness function 22

23 Solution? Run as a parallel application Call the hill climbing routine every N generations with N>1 The hope is: Calling hill climbing routine every N generations speeds up routine Calling hill climbing routine every N generations will not greatly degrade the quality of the solution Hope was born out by experiment Improved Mansour-Fox algorithm reduced the run time of the adaptive grid simulation 23

24 What was learned to this point? What sine qua nons must be observed to make GA effective Use domain specific knowledge Improve the algorithm of the GA Important specifics: Ignore communication between split and unsplit cells in fitness function Use improved Mansour-Fox algorithm 24

25 Application of heuristics Problem run on 16 nodes of SP1 and SP2 Initial grid is 256 x 256 Multiple levels of adaptation Nonuniform distribution of mass 2 Pressure cycle= 0 t=0 pre 3.3 Nonuniform distribution of energy y x

26 Initial results using normal Mansour-Fox algorithm Parallel run time without GA = 1115 seconds Balance load only run time = 1240 seconds Mansour-Fox GA parameters: Population size = 320 Generations = 100 Adjusted weighting for cost of communication Weight Wall time Speed up GA Return Balance Comm. Time Run time with Mansour-Fox algorithm = 928 seconds with 44 seconds GA run time

27 Ran using the improved Mansour-Fox algorithm Allow the hill climbing routine to be called less often Tuned the GA Adjust to find optimum: Population size Number of generations Frequency of hill climbing routine in improved Mansour-Fox algorithm 27

28 Results with GA population size = 480 Often Size Gen Wall time Speed up GA Return Balance Comm. Time

29 Results with GA population size = 320 Often Size Gen Wall time Speed up GA Return Balance Comm. Time

30 Results with GA population size = 240 Often Size Gen Wall time Speed up GA Return Balance Comm. Time

31 Summary of results GA run time was in the range 18 to 117 seconds Simulation time was in the range of 1004 to 886 seconds Communication time was in the range of 283 to 325 seconds Run time reduced from 1115 seconds to 886 seconds by using the improved Mansour-Fox algorithm Speed up = 26% Population size = 320 Generations for GA =100 Frequency of calling hill climbing routine = 4 31

32 Improved Mansour-Fox algorithm By calling the hill climbing routine less often: GA run time is reduced from about 117 to about 2O seconds The algorithm still returned a good solution. Best run times with frequency of 4, longest with frequency of 1 Simulation run time (seconds) Generations=200 Generations=150 Generations= every 1 every 2 every 3 every 4 every 5 Frequency of hill climbing routine 32

33 Use of tuning the Genetic Algorithm Apply to other architectures Apply to other problems 33

34 Tuning for SP1 applied to SP2 Ran same simulation on the MHPCC SP2 Used same GA control parameters except: Conjecture is that we need a smaller weighting for communication down from Wall time Speed up GA Return Balance Comm. Time Weight Using tuning data from SP1 runs enabled 75% speed up on SP2 Conjecture is shown to be true Showed that tuning for one machine can be applied to another 34

35 Did comparison to bisection methods The Hybrid GA with improved Mansour-Fox algorithm produces the best results Machine Algorithm Run time Comm. Time Balance SP1 Best GA SP1 Centroid Bisection SP1 Coordinate Bisection SP2 Best GA SP2 Centroid Bisection SP2 Coordinate Bisection

36 Application to a different problem on SP1 Useful for performing trade studies Similar simulation Uniform blobs of energy and mass Deposited at different times Used best GA control parameters from the previous simulation (Frequency = 4, Size = 320, Generations = 100) Algorithm Wall time Speed up GA Return Balance Comm. Time Baseline Bisection GA Enabled 14% improvement in run time Recursive bisection caused a slow down 36

37 Future directions Using the GA Different architectures Have done simulations of SP2 SMP nodes 3d problems Oil field studies Improve the adaptive grid framework 3d Polygonal cells Edge based scheme 37

38 Summary Developed frameworks for Parallel adaptive mesh program Parallel Genetic Algorithm Studied the use of Genetic Algorithm to perform dynamic allocation of grids to processors Discovered two sine qua nons to enable the GA to be effective Use domain specific knowledge Improve the algorithm of the GA Showed the hybrid GA to be effective Using: Improved the Mansour-Fox algorithm Domain Knowledge of propagating shock waves Improved run time of an example calculation 75% 38

TIMOTHY HAROLD KAISER

TIMOTHY HAROLD KAISER Dynamic Load Distributions for Adaptive Computations on MIMD Machines using hybrid Genetic Algorithms by TIMOTHY HAROLD KAISER Doctor of Philosophy, Computer Science, The University of New Mexico, 1997

More information

Example 13 - Shock Tube

Example 13 - Shock Tube Example 13 - Shock Tube Summary This famous experiment is interesting for observing the shock-wave propagation. Moreover, this case uses the representation of perfect gas and compares the different formulations:

More information

Introduction to Genetic Algorithms. Based on Chapter 10 of Marsland Chapter 9 of Mitchell

Introduction to Genetic Algorithms. Based on Chapter 10 of Marsland Chapter 9 of Mitchell Introduction to Genetic Algorithms Based on Chapter 10 of Marsland Chapter 9 of Mitchell Genetic Algorithms - History Pioneered by John Holland in the 1970s Became popular in the late 1980s Based on ideas

More information

Research Incubator: Combinatorial Optimization. Dr. Lixin Tao December 9, 2003

Research Incubator: Combinatorial Optimization. Dr. Lixin Tao December 9, 2003 Research Incubator: Combinatorial Optimization Dr. Lixin Tao December 9, 23 Content General Nature of Research on Combinatorial Optimization Problem Identification and Abstraction Problem Properties and

More information

Data Partitioning. Figure 1-31: Communication Topologies. Regular Partitions

Data Partitioning. Figure 1-31: Communication Topologies. Regular Partitions Data In single-program multiple-data (SPMD) parallel programs, global data is partitioned, with a portion of the data assigned to each processing node. Issues relevant to choosing a partitioning strategy

More information

Adaptive-Mesh-Refinement Pattern

Adaptive-Mesh-Refinement Pattern Adaptive-Mesh-Refinement Pattern I. Problem Data-parallelism is exposed on a geometric mesh structure (either irregular or regular), where each point iteratively communicates with nearby neighboring points

More information

Multigrid Pattern. I. Problem. II. Driving Forces. III. Solution

Multigrid Pattern. I. Problem. II. Driving Forces. III. Solution Multigrid Pattern I. Problem Problem domain is decomposed into a set of geometric grids, where each element participates in a local computation followed by data exchanges with adjacent neighbors. The grids

More information

Artificial Intelligence

Artificial Intelligence Artificial Intelligence Informed Search and Exploration Chapter 4 (4.3 4.6) Searching: So Far We ve discussed how to build goal-based and utility-based agents that search to solve problems We ve also presented

More information

RAMSES on the GPU: An OpenACC-Based Approach

RAMSES on the GPU: An OpenACC-Based Approach RAMSES on the GPU: An OpenACC-Based Approach Claudio Gheller (ETHZ-CSCS) Giacomo Rosilho de Souza (EPFL Lausanne) Romain Teyssier (University of Zurich) Markus Wetzstein (ETHZ-CSCS) PRACE-2IP project EU

More information

Genetic Algorithm for Circuit Partitioning

Genetic Algorithm for Circuit Partitioning Genetic Algorithm for Circuit Partitioning ZOLTAN BARUCH, OCTAVIAN CREŢ, KALMAN PUSZTAI Computer Science Department, Technical University of Cluj-Napoca, 26, Bariţiu St., 3400 Cluj-Napoca, Romania {Zoltan.Baruch,

More information

Local Search (Ch )

Local Search (Ch ) Local Search (Ch. 4-4.1) Local search Before we tried to find a path from the start state to a goal state using a fringe set Now we will look at algorithms that do not care about a fringe, but just neighbors

More information

Solving Partial Differential Equations on Overlapping Grids

Solving Partial Differential Equations on Overlapping Grids **FULL TITLE** ASP Conference Series, Vol. **VOLUME**, **YEAR OF PUBLICATION** **NAMES OF EDITORS** Solving Partial Differential Equations on Overlapping Grids William D. Henshaw Centre for Applied Scientific

More information

Outline. Best-first search. Greedy best-first search A* search Heuristics Local search algorithms

Outline. Best-first search. Greedy best-first search A* search Heuristics Local search algorithms Outline Best-first search Greedy best-first search A* search Heuristics Local search algorithms Hill-climbing search Beam search Simulated annealing search Genetic algorithms Constraint Satisfaction Problems

More information

Implementation of a 3D Hilbert SFC into a Parallel Cartesian- Grid Flow Solver

Implementation of a 3D Hilbert SFC into a Parallel Cartesian- Grid Flow Solver Implementation of a 3D Hilbert SFC into a Parallel Cartesian- Grid Flow Solver Stephen M. Ruffin and Jinwook Lee* Abstract - The efficient parallel computation of unstructured grid flow solver requires

More information

Advanced Search Genetic algorithm

Advanced Search Genetic algorithm Advanced Search Genetic algorithm Yingyu Liang yliang@cs.wisc.edu Computer Sciences Department University of Wisconsin, Madison [Based on slides from Jerry Zhu, Andrew Moore http://www.cs.cmu.edu/~awm/tutorials

More information

Week 3: MPI. Day 04 :: Domain decomposition, load balancing, hybrid particlemesh

Week 3: MPI. Day 04 :: Domain decomposition, load balancing, hybrid particlemesh Week 3: MPI Day 04 :: Domain decomposition, load balancing, hybrid particlemesh methods Domain decompositon Goals of parallel computing Solve a bigger problem Operate on more data (grid points, particles,

More information

A PARALLEL ALGORITHM FOR THE DEFORMATION AND INTERACTION OF STRUCTURES MODELED WITH LAGRANGE MESHES IN AUTODYN-3D

A PARALLEL ALGORITHM FOR THE DEFORMATION AND INTERACTION OF STRUCTURES MODELED WITH LAGRANGE MESHES IN AUTODYN-3D 3 rd International Symposium on Impact Engineering 98, 7-9 December 1998, Singapore A PARALLEL ALGORITHM FOR THE DEFORMATION AND INTERACTION OF STRUCTURES MODELED WITH LAGRANGE MESHES IN AUTODYN-3D M.

More information

Parallel Computing. Parallel Algorithm Design

Parallel Computing. Parallel Algorithm Design Parallel Computing Parallel Algorithm Design Task/Channel Model Parallel computation = set of tasks Task Program Local memory Collection of I/O ports Tasks interact by sending messages through channels

More information

Graph Partitioning for High-Performance Scientific Simulations. Advanced Topics Spring 2008 Prof. Robert van Engelen

Graph Partitioning for High-Performance Scientific Simulations. Advanced Topics Spring 2008 Prof. Robert van Engelen Graph Partitioning for High-Performance Scientific Simulations Advanced Topics Spring 2008 Prof. Robert van Engelen Overview Challenges for irregular meshes Modeling mesh-based computations as graphs Static

More information

ALE Adaptive Mesh Refinement in LS-DYNA

ALE Adaptive Mesh Refinement in LS-DYNA 12 th International LS-DYNA Users Conference FSI/ALE(2) ALE Adaptive Mesh Refinement in LS-DYNA Nicolas AQUELET Livermore Software Technology Corp. 7374 Las Positas Rd Livermore CA94550 aquelet@lstc.com

More information

Coupling of STAR-CCM+ to Other Theoretical or Numerical Solutions. Milovan Perić

Coupling of STAR-CCM+ to Other Theoretical or Numerical Solutions. Milovan Perić Coupling of STAR-CCM+ to Other Theoretical or Numerical Solutions Milovan Perić Contents The need to couple STAR-CCM+ with other theoretical or numerical solutions Coupling approaches: surface and volume

More information

Parallel Algorithms: Adaptive Mesh Refinement (AMR) method and its implementation

Parallel Algorithms: Adaptive Mesh Refinement (AMR) method and its implementation Parallel Algorithms: Adaptive Mesh Refinement (AMR) method and its implementation Massimiliano Guarrasi m.guarrasi@cineca.it Super Computing Applications and Innovation Department AMR - Introduction Solving

More information

Store Separation Simulation using Oct-tree Grid Based Solver

Store Separation Simulation using Oct-tree Grid Based Solver SAROD 2009 142 Symposium on Applied Aerodynamics and Design of Aerospace Vehicles (SAROD 2009) December 10-12, 2009, Bengaluru, India Store Separation Simulation using Oct-tree Grid Based Solver Saurabh

More information

Uninformed Search Methods. Informed Search Methods. Midterm Exam 3/13/18. Thursday, March 15, 7:30 9:30 p.m. room 125 Ag Hall

Uninformed Search Methods. Informed Search Methods. Midterm Exam 3/13/18. Thursday, March 15, 7:30 9:30 p.m. room 125 Ag Hall Midterm Exam Thursday, March 15, 7:30 9:30 p.m. room 125 Ag Hall Covers topics through Decision Trees and Random Forests (does not include constraint satisfaction) Closed book 8.5 x 11 sheet with notes

More information

Parallel Algorithm Design. Parallel Algorithm Design p. 1

Parallel Algorithm Design. Parallel Algorithm Design p. 1 Parallel Algorithm Design Parallel Algorithm Design p. 1 Overview Chapter 3 from Michael J. Quinn, Parallel Programming in C with MPI and OpenMP Another resource: http://www.mcs.anl.gov/ itf/dbpp/text/node14.html

More information

CHAPTER 2 CONVENTIONAL AND NON-CONVENTIONAL TECHNIQUES TO SOLVE ORPD PROBLEM

CHAPTER 2 CONVENTIONAL AND NON-CONVENTIONAL TECHNIQUES TO SOLVE ORPD PROBLEM 20 CHAPTER 2 CONVENTIONAL AND NON-CONVENTIONAL TECHNIQUES TO SOLVE ORPD PROBLEM 2.1 CLASSIFICATION OF CONVENTIONAL TECHNIQUES Classical optimization methods can be classified into two distinct groups:

More information

Radio Network Planning with Combinatorial Optimisation Algorithms

Radio Network Planning with Combinatorial Optimisation Algorithms Author manuscript, published in "ACTS Mobile Telecommunications Summit 96, Granada : Spain (1996)" Radio Network Planning with Combinatorial Optimisation Algorithms P. Calégari, F. Guidec, P. Kuonen, EPFL,

More information

Multigrid Algorithms for Three-Dimensional RANS Calculations - The SUmb Solver

Multigrid Algorithms for Three-Dimensional RANS Calculations - The SUmb Solver Multigrid Algorithms for Three-Dimensional RANS Calculations - The SUmb Solver Juan J. Alonso Department of Aeronautics & Astronautics Stanford University CME342 Lecture 14 May 26, 2014 Outline Non-linear

More information

Adaptive Mesh Refinement

Adaptive Mesh Refinement Aleander Knebe, Universidad Autonoma de Madrid Adaptive Mesh Refinement AMR codes Poisson s equation ΔΦ( ) = 4πGρ( ) Poisson s equation F ( ) = m Φ( ) ΔΦ( ) = 4πGρ( ) particle approach F ( Gm i ) = i m

More information

Realtime Water Simulation on GPU. Nuttapong Chentanez NVIDIA Research

Realtime Water Simulation on GPU. Nuttapong Chentanez NVIDIA Research 1 Realtime Water Simulation on GPU Nuttapong Chentanez NVIDIA Research 2 3 Overview Approaches to realtime water simulation Hybrid shallow water solver + particles Hybrid 3D tall cell water solver + particles

More information

Continuum-Microscopic Models

Continuum-Microscopic Models Scientific Computing and Numerical Analysis Seminar October 1, 2010 Outline Heterogeneous Multiscale Method Adaptive Mesh ad Algorithm Refinement Equation-Free Method Incorporates two scales (length, time

More information

Development of an Integrated Computational Simulation Method for Fluid Driven Structure Movement and Acoustics

Development of an Integrated Computational Simulation Method for Fluid Driven Structure Movement and Acoustics Development of an Integrated Computational Simulation Method for Fluid Driven Structure Movement and Acoustics I. Pantle Fachgebiet Strömungsmaschinen Karlsruher Institut für Technologie KIT Motivation

More information

Using Genetic Algorithms to solve the Minimum Labeling Spanning Tree Problem

Using Genetic Algorithms to solve the Minimum Labeling Spanning Tree Problem Using to solve the Minimum Labeling Spanning Tree Problem Final Presentation, oliverr@umd.edu Advisor: Dr Bruce L. Golden, bgolden@rhsmith.umd.edu R. H. Smith School of Business (UMD) May 3, 2012 1 / 42

More information

Applications. Oversampled 3D scan data. ~150k triangles ~80k triangles

Applications. Oversampled 3D scan data. ~150k triangles ~80k triangles Mesh Simplification Applications Oversampled 3D scan data ~150k triangles ~80k triangles 2 Applications Overtessellation: E.g. iso-surface extraction 3 Applications Multi-resolution hierarchies for efficient

More information

Debojyoti Ghosh. Adviser: Dr. James Baeder Alfred Gessow Rotorcraft Center Department of Aerospace Engineering

Debojyoti Ghosh. Adviser: Dr. James Baeder Alfred Gessow Rotorcraft Center Department of Aerospace Engineering Debojyoti Ghosh Adviser: Dr. James Baeder Alfred Gessow Rotorcraft Center Department of Aerospace Engineering To study the Dynamic Stalling of rotor blade cross-sections Unsteady Aerodynamics: Time varying

More information

N-Queens problem. Administrative. Local Search

N-Queens problem. Administrative. Local Search Local Search CS151 David Kauchak Fall 2010 http://www.youtube.com/watch?v=4pcl6-mjrnk Some material borrowed from: Sara Owsley Sood and others Administrative N-Queens problem Assign 1 grading Assign 2

More information

Accurate and Efficient Turbomachinery Simulation. Chad Custer, PhD Turbomachinery Technical Specialist

Accurate and Efficient Turbomachinery Simulation. Chad Custer, PhD Turbomachinery Technical Specialist Accurate and Efficient Turbomachinery Simulation Chad Custer, PhD Turbomachinery Technical Specialist Outline Turbomachinery simulation advantages Axial fan optimization Description of design objectives

More information

A Two-Stage Stochastic Programming Approach for Location-Allocation Models in Uncertain Environments

A Two-Stage Stochastic Programming Approach for Location-Allocation Models in Uncertain Environments A Two-Stage Stochastic Programming Approach for Location-Allocation in Uncertain Environments Markus Kaiser, Kathrin Klamroth Optimization & Approximation Department of Mathematics University of Wuppertal

More information

Hybridization EVOLUTIONARY COMPUTING. Reasons for Hybridization - 1. Naming. Reasons for Hybridization - 3. Reasons for Hybridization - 2

Hybridization EVOLUTIONARY COMPUTING. Reasons for Hybridization - 1. Naming. Reasons for Hybridization - 3. Reasons for Hybridization - 2 Hybridization EVOLUTIONARY COMPUTING Hybrid Evolutionary Algorithms hybridization of an EA with local search techniques (commonly called memetic algorithms) EA+LS=MA constructive heuristics exact methods

More information

Unstructured 3-D Grid Partitioning Methods Based On Genetic Algorithms

Unstructured 3-D Grid Partitioning Methods Based On Genetic Algorithms Unstructured 3-D Grid Partitioning Methods Based On Genetic Algorithms K.C. Giannakoglou and A.P. Giotis 1 Abstract. In this paper, two methods that are capable to efficiently partition 3-D unstructured

More information

Lecture 4: Principles of Parallel Algorithm Design (part 4)

Lecture 4: Principles of Parallel Algorithm Design (part 4) Lecture 4: Principles of Parallel Algorithm Design (part 4) 1 Mapping Technique for Load Balancing Minimize execution time Reduce overheads of execution Sources of overheads: Inter-process interaction

More information

Genetic Programming: A study on Computer Language

Genetic Programming: A study on Computer Language Genetic Programming: A study on Computer Language Nilam Choudhary Prof.(Dr.) Baldev Singh Er. Gaurav Bagaria Abstract- this paper describes genetic programming in more depth, assuming that the reader is

More information

Shallow Water Simulations on Graphics Hardware

Shallow Water Simulations on Graphics Hardware Shallow Water Simulations on Graphics Hardware Ph.D. Thesis Presentation 2014-06-27 Martin Lilleeng Sætra Outline Introduction Parallel Computing and the GPU Simulating Shallow Water Flow Topics of Thesis

More information

A NOVEL APPROACH FOR PRIORTIZATION OF OPTIMIZED TEST CASES

A NOVEL APPROACH FOR PRIORTIZATION OF OPTIMIZED TEST CASES A NOVEL APPROACH FOR PRIORTIZATION OF OPTIMIZED TEST CASES Abhishek Singhal Amity School of Engineering and Technology Amity University Noida, India asinghal1@amity.edu Swati Chandna Amity School of Engineering

More information

Interconnection Networks: Topology. Prof. Natalie Enright Jerger

Interconnection Networks: Topology. Prof. Natalie Enright Jerger Interconnection Networks: Topology Prof. Natalie Enright Jerger Topology Overview Definition: determines arrangement of channels and nodes in network Analogous to road map Often first step in network design

More information

Numerical Methods for PDEs. SSC Workgroup Meetings Juan J. Alonso October 8, SSC Working Group Meetings, JJA 1

Numerical Methods for PDEs. SSC Workgroup Meetings Juan J. Alonso October 8, SSC Working Group Meetings, JJA 1 Numerical Methods for PDEs SSC Workgroup Meetings Juan J. Alonso October 8, 2001 SSC Working Group Meetings, JJA 1 Overview These notes are meant to be an overview of the various memory access patterns

More information

Evolutionary Algorithms: Perfecting the Art of Good Enough. Liz Sander

Evolutionary Algorithms: Perfecting the Art of Good Enough. Liz Sander Evolutionary Algorithms: Perfecting the Art of Good Enough Liz Sander Source: wikipedia.org Source: fishbase.org Source: youtube.com Sometimes, we can t find the best solution. Sometimes, we can t find

More information

B553 Lecture 12: Global Optimization

B553 Lecture 12: Global Optimization B553 Lecture 12: Global Optimization Kris Hauser February 20, 2012 Most of the techniques we have examined in prior lectures only deal with local optimization, so that we can only guarantee convergence

More information

Escaping Local Optima: Genetic Algorithm

Escaping Local Optima: Genetic Algorithm Artificial Intelligence Escaping Local Optima: Genetic Algorithm Dae-Won Kim School of Computer Science & Engineering Chung-Ang University We re trying to escape local optima To achieve this, we have learned

More information

CFD Project Workflow Guide

CFD Project Workflow Guide CFD Project Workflow Guide Contents Select a problem with known results for proof-of-concept testing... 1 Set up and run a coarse test case... 2 Select and calibrate numerical methods... 3 Minimize & quantify

More information

CHAPTER 4 GENETIC ALGORITHM

CHAPTER 4 GENETIC ALGORITHM 69 CHAPTER 4 GENETIC ALGORITHM 4.1 INTRODUCTION Genetic Algorithms (GAs) were first proposed by John Holland (Holland 1975) whose ideas were applied and expanded on by Goldberg (Goldberg 1989). GAs is

More information

ANSYS AIM Tutorial Compressible Flow in a Nozzle

ANSYS AIM Tutorial Compressible Flow in a Nozzle ANSYS AIM Tutorial Compressible Flow in a Nozzle Author(s): Sebastian Vecchi Created using ANSYS AIM 18.1 Problem Specification Pre-Analysis & Start Up Pre-Analysis Start-Up Geometry Import Geometry Mesh

More information

Multiprocessors and Thread Level Parallelism Chapter 4, Appendix H CS448. The Greed for Speed

Multiprocessors and Thread Level Parallelism Chapter 4, Appendix H CS448. The Greed for Speed Multiprocessors and Thread Level Parallelism Chapter 4, Appendix H CS448 1 The Greed for Speed Two general approaches to making computers faster Faster uniprocessor All the techniques we ve been looking

More information

6. Parallel Volume Rendering Algorithms

6. Parallel Volume Rendering Algorithms 6. Parallel Volume Algorithms This chapter introduces a taxonomy of parallel volume rendering algorithms. In the thesis statement we claim that parallel algorithms may be described by "... how the tasks

More information

Parallel Traveling Salesman. PhD Student: Viet Anh Trinh Advisor: Professor Feng Gu.

Parallel Traveling Salesman. PhD Student: Viet Anh Trinh Advisor: Professor Feng Gu. Parallel Traveling Salesman PhD Student: Viet Anh Trinh Advisor: Professor Feng Gu Agenda 1. Traveling salesman introduction 2. Genetic Algorithm for TSP 3. Tree Search for TSP Travelling Salesman - Set

More information

Genetic Algorithms Variations and Implementation Issues

Genetic Algorithms Variations and Implementation Issues Genetic Algorithms Variations and Implementation Issues CS 431 Advanced Topics in AI Classic Genetic Algorithms GAs as proposed by Holland had the following properties: Randomly generated population Binary

More information

Administrative. Local Search!

Administrative. Local Search! Administrative Local Search! CS311 David Kauchak Spring 2013 Assignment 2 due Tuesday before class Written problems 2 posted Class participation http://www.youtube.com/watch? v=irhfvdphfzq&list=uucdoqrpqlqkvctckzqa

More information

Programming as Successive Refinement. Partitioning for Performance

Programming as Successive Refinement. Partitioning for Performance Programming as Successive Refinement Not all issues dealt with up front Partitioning often independent of architecture, and done first View machine as a collection of communicating processors balancing

More information

Cost Effective Gas Turbine Simulation

Cost Effective Gas Turbine Simulation Cost Effective Gas Turbine Simulation Defence Education Conference - Athens GR - 1 June 2017 War Museum Ioannis Templalexis Associate Professor (elected) Hellenic Air Force Academy Department of Aeronautical

More information

CS 378: Computer Game Technology

CS 378: Computer Game Technology CS 378: Computer Game Technology Dynamic Path Planning, Flocking Spring 2012 University of Texas at Austin CS 378 Game Technology Don Fussell Dynamic Path Planning! What happens when the environment changes

More information

WAVE PATTERNS, WAVE INDUCED FORCES AND MOMENTS FOR A GRAVITY BASED STRUCTURE PREDICTED USING CFD

WAVE PATTERNS, WAVE INDUCED FORCES AND MOMENTS FOR A GRAVITY BASED STRUCTURE PREDICTED USING CFD Proceedings of the ASME 2011 30th International Conference on Ocean, Offshore and Arctic Engineering OMAE2011 June 19-24, 2011, Rotterdam, The Netherlands OMAE2011-49593 WAVE PATTERNS, WAVE INDUCED FORCES

More information

Genetic programming. Lecture Genetic Programming. LISP as a GP language. LISP structure. S-expressions

Genetic programming. Lecture Genetic Programming. LISP as a GP language. LISP structure. S-expressions Genetic programming Lecture Genetic Programming CIS 412 Artificial Intelligence Umass, Dartmouth One of the central problems in computer science is how to make computers solve problems without being explicitly

More information

A Comparison of Load Balancing Algorithms for Parallel Computations

A Comparison of Load Balancing Algorithms for Parallel Computations Syracuse University SURFACE Electrical Engineering and Computer Science Technical Reports College of Engineering and Computer Science 9-1991 A Comparison of Load Balancing Algorithms for Parallel Computations

More information

Dynamic Load Partitioning Strategies for Managing Data of Space and Time Heterogeneity in Parallel SAMR Applications

Dynamic Load Partitioning Strategies for Managing Data of Space and Time Heterogeneity in Parallel SAMR Applications Dynamic Load Partitioning Strategies for Managing Data of Space and Time Heterogeneity in Parallel SAMR Applications Xiaolin Li and Manish Parashar The Applied Software Systems Laboratory Department of

More information

Challenges and recent progress in developing numerical methods for multi-material ALE Hydrocodes

Challenges and recent progress in developing numerical methods for multi-material ALE Hydrocodes Challenges and recent progress in developing numerical methods for multi-material ALE Hydrocodes ICFD 25 year Anniversary Conference 15-16 th September 2008 Oxford University Andrew Barlow AWE Introduction

More information

GENETIC ALGORITHM with Hands-On exercise

GENETIC ALGORITHM with Hands-On exercise GENETIC ALGORITHM with Hands-On exercise Adopted From Lecture by Michael Negnevitsky, Electrical Engineering & Computer Science University of Tasmania 1 Objective To understand the processes ie. GAs Basic

More information

Variational Methods II

Variational Methods II Mathematical Foundations of Computer Graphics and Vision Variational Methods II Luca Ballan Institute of Visual Computing Last Lecture If we have a topological vector space with an inner product and functionals

More information

A Genetic Algorithm for Minimum Tetrahedralization of a Convex Polyhedron

A Genetic Algorithm for Minimum Tetrahedralization of a Convex Polyhedron A Genetic Algorithm for Minimum Tetrahedralization of a Convex Polyhedron Kiat-Choong Chen Ian Hsieh Cao An Wang Abstract A minimum tetrahedralization of a convex polyhedron is a partition of the convex

More information

Sci.Int.(Lahore),28(1), ,2016 ISSN ; CODEN: SINTE 8 201

Sci.Int.(Lahore),28(1), ,2016 ISSN ; CODEN: SINTE 8 201 Sci.Int.(Lahore),28(1),201-209,2016 ISSN 1013-5316; CODEN: SINTE 8 201 A NOVEL PLANT PROPAGATION ALGORITHM: MODIFICATIONS AND IMPLEMENTATION Muhammad Sulaiman 1, Abdel Salhi 2, Eric S Fraga 3, Wali Khan

More information

Parallel Mesh Multiplication for Code_Saturne

Parallel Mesh Multiplication for Code_Saturne Parallel Mesh Multiplication for Code_Saturne Pavla Kabelikova, Ales Ronovsky, Vit Vondrak a Dept. of Applied Mathematics, VSB-Technical University of Ostrava, Tr. 17. listopadu 15, 708 00 Ostrava, Czech

More information

CutLeader Nesting Technology

CutLeader Nesting Technology CutLeader Technology algorithm is the soul of nesting software. For example knapsack algorithm, Pair technology, are able to get a better nesting result. The former is the approximate optimization algorithm;

More information

CFD Best Practice Guidelines: A process to understand CFD results and establish Simulation versus Reality

CFD Best Practice Guidelines: A process to understand CFD results and establish Simulation versus Reality CFD Best Practice Guidelines: A process to understand CFD results and establish Simulation versus Reality Judd Kaiser ANSYS Inc. judd.kaiser@ansys.com 2005 ANSYS, Inc. 1 ANSYS, Inc. Proprietary Overview

More information

Adaptive-Mesh-Refinement Hydrodynamic GPU Computation in Astrophysics

Adaptive-Mesh-Refinement Hydrodynamic GPU Computation in Astrophysics Adaptive-Mesh-Refinement Hydrodynamic GPU Computation in Astrophysics H. Y. Schive ( 薛熙于 ) Graduate Institute of Physics, National Taiwan University Leung Center for Cosmology and Particle Astrophysics

More information

Object Oriented Simulation of Multiphase Flow

Object Oriented Simulation of Multiphase Flow Object Oriented Simulation of Multiphase Flow P.Klebert and O.J.Nydal Norwegian University of Science and Technology (NTNU) 7491 Trondheim Norway Abstract Pipelines in offshore oil and gas production systems

More information

2.7 Cloth Animation. Jacobs University Visualization and Computer Graphics Lab : Advanced Graphics - Chapter 2 123

2.7 Cloth Animation. Jacobs University Visualization and Computer Graphics Lab : Advanced Graphics - Chapter 2 123 2.7 Cloth Animation 320491: Advanced Graphics - Chapter 2 123 Example: Cloth draping Image Michael Kass 320491: Advanced Graphics - Chapter 2 124 Cloth using mass-spring model Network of masses and springs

More information

Direct Numerical Simulation of a Low Pressure Turbine Cascade. Christoph Müller

Direct Numerical Simulation of a Low Pressure Turbine Cascade. Christoph Müller Low Pressure NOFUN 2015, Braunschweig, Overview PostProcessing Experimental test facility Grid generation Inflow turbulence Conclusion and slide 2 / 16 Project Scale resolving Simulations give insight

More information

Parallel Numerics, WT 2013/ Introduction

Parallel Numerics, WT 2013/ Introduction Parallel Numerics, WT 2013/2014 1 Introduction page 1 of 122 Scope Revise standard numerical methods considering parallel computations! Required knowledge Numerics Parallel Programming Graphs Literature

More information

Seminar on. A Coarse-Grain Parallel Formulation of Multilevel k-way Graph Partitioning Algorithm

Seminar on. A Coarse-Grain Parallel Formulation of Multilevel k-way Graph Partitioning Algorithm Seminar on A Coarse-Grain Parallel Formulation of Multilevel k-way Graph Partitioning Algorithm Mohammad Iftakher Uddin & Mohammad Mahfuzur Rahman Matrikel Nr: 9003357 Matrikel Nr : 9003358 Masters of

More information

Artificial Intelligence Application (Genetic Algorithm)

Artificial Intelligence Application (Genetic Algorithm) Babylon University College of Information Technology Software Department Artificial Intelligence Application (Genetic Algorithm) By Dr. Asaad Sabah Hadi 2014-2015 EVOLUTIONARY ALGORITHM The main idea about

More information

Scalability Study of Particle Method with Dynamic Load Balancing

Scalability Study of Particle Method with Dynamic Load Balancing Scalability Study of Particle Method with Dynamic Load Balancing Hailong Teng Livermore Software Technology Corp. Abstract We introduce an efficient load-balancing algorithm for particle method (Particle

More information

Path Planning. Marcello Restelli. Dipartimento di Elettronica e Informazione Politecnico di Milano tel:

Path Planning. Marcello Restelli. Dipartimento di Elettronica e Informazione Politecnico di Milano   tel: Marcello Restelli Dipartimento di Elettronica e Informazione Politecnico di Milano email: restelli@elet.polimi.it tel: 02 2399 3470 Path Planning Robotica for Computer Engineering students A.A. 2006/2007

More information

Dynamic Spatial Partitioning for Real-Time Visibility Determination. Joshua Shagam Computer Science

Dynamic Spatial Partitioning for Real-Time Visibility Determination. Joshua Shagam Computer Science Dynamic Spatial Partitioning for Real-Time Visibility Determination Joshua Shagam Computer Science Master s Defense May 2, 2003 Problem Complex 3D environments have large numbers of objects Computer hardware

More information

A Modified Genetic Algorithm for Process Scheduling in Distributed System

A Modified Genetic Algorithm for Process Scheduling in Distributed System A Modified Genetic Algorithm for Process Scheduling in Distributed System Vinay Harsora B.V.M. Engineering College Charatar Vidya Mandal Vallabh Vidyanagar, India Dr.Apurva Shah G.H.Patel College of Engineering

More information

Genetic Algorithm Performance with Different Selection Methods in Solving Multi-Objective Network Design Problem

Genetic Algorithm Performance with Different Selection Methods in Solving Multi-Objective Network Design Problem etic Algorithm Performance with Different Selection Methods in Solving Multi-Objective Network Design Problem R. O. Oladele Department of Computer Science University of Ilorin P.M.B. 1515, Ilorin, NIGERIA

More information

The Automatic Design of Batch Processing Systems

The Automatic Design of Batch Processing Systems The Automatic Design of Batch Processing Systems by Barry Dwyer, M.A., D.A.E., Grad.Dip. A thesis submitted for the degree of Doctor of Philosophy in the Department of Computer Science University of Adelaide

More information

Introduction to Genetic Algorithms. Genetic Algorithms

Introduction to Genetic Algorithms. Genetic Algorithms Introduction to Genetic Algorithms Genetic Algorithms We ve covered enough material that we can write programs that use genetic algorithms! More advanced example of using arrays Could be better written

More information

Using Genetic Algorithms to optimize ACS-TSP

Using Genetic Algorithms to optimize ACS-TSP Using Genetic Algorithms to optimize ACS-TSP Marcin L. Pilat and Tony White School of Computer Science, Carleton University, 1125 Colonel By Drive, Ottawa, ON, K1S 5B6, Canada {mpilat,arpwhite}@scs.carleton.ca

More information

Spatial Data Structures for Computer Graphics

Spatial Data Structures for Computer Graphics Spatial Data Structures for Computer Graphics Page 1 of 65 http://www.cse.iitb.ac.in/ sharat November 2008 Spatial Data Structures for Computer Graphics Page 1 of 65 http://www.cse.iitb.ac.in/ sharat November

More information

University of Florida CISE department Gator Engineering. Clustering Part 4

University of Florida CISE department Gator Engineering. Clustering Part 4 Clustering Part 4 Dr. Sanjay Ranka Professor Computer and Information Science and Engineering University of Florida, Gainesville DBSCAN DBSCAN is a density based clustering algorithm Density = number of

More information

Introduction to Optimization

Introduction to Optimization Introduction to Optimization Approximation Algorithms and Heuristics November 21, 2016 École Centrale Paris, Châtenay-Malabry, France Dimo Brockhoff Inria Saclay Ile-de-France 2 Exercise: The Knapsack

More information

Job Shop Scheduling Problem (JSSP) Genetic Algorithms Critical Block and DG distance Neighbourhood Search

Job Shop Scheduling Problem (JSSP) Genetic Algorithms Critical Block and DG distance Neighbourhood Search A JOB-SHOP SCHEDULING PROBLEM (JSSP) USING GENETIC ALGORITHM (GA) Mahanim Omar, Adam Baharum, Yahya Abu Hasan School of Mathematical Sciences, Universiti Sains Malaysia 11800 Penang, Malaysia Tel: (+)

More information

Efficient Storage and Processing of Adaptive Triangular Grids using Sierpinski Curves

Efficient Storage and Processing of Adaptive Triangular Grids using Sierpinski Curves Efficient Storage and Processing of Adaptive Triangular Grids using Sierpinski Curves Csaba Attila Vigh, Dr. Michael Bader Department of Informatics, TU München JASS 2006, course 2: Numerical Simulation:

More information

Evolutionary Computation Algorithms for Cryptanalysis: A Study

Evolutionary Computation Algorithms for Cryptanalysis: A Study Evolutionary Computation Algorithms for Cryptanalysis: A Study Poonam Garg Information Technology and Management Dept. Institute of Management Technology Ghaziabad, India pgarg@imt.edu Abstract The cryptanalysis

More information

More Animation Techniques

More Animation Techniques CS 231 More Animation Techniques So much more Animation Procedural animation Particle systems Free-form deformation Natural Phenomena 1 Procedural Animation Rule based animation that changes/evolves over

More information

A STUDY OF LOAD IMBALANCE FOR PARALLEL RESERVOIR SIMULATION WITH MULTIPLE PARTITIONING STRATEGIES. A Thesis XUYANG GUO

A STUDY OF LOAD IMBALANCE FOR PARALLEL RESERVOIR SIMULATION WITH MULTIPLE PARTITIONING STRATEGIES. A Thesis XUYANG GUO A STUDY OF LOAD IMBALANCE FOR PARALLEL RESERVOIR SIMULATION WITH MULTIPLE PARTITIONING STRATEGIES A Thesis by XUYANG GUO Submitted to the Office of Graduate and Professional Studies of Texas A&M University

More information

AIRFOIL SHAPE OPTIMIZATION USING IMPROVED SIMPLE GENETIC ALGORITHM (ISGA)

AIRFOIL SHAPE OPTIMIZATION USING IMPROVED SIMPLE GENETIC ALGORITHM (ISGA) HEFAT007 5 th International onference on Heat Transfer, Fluid Mechanics and Thermodynamics Sun ity, South Africa MK1 AIRFOIL SHAPE OPTIMIZATION USING IMPROVED SIMPLE GENETI ALGORITHM (ISGA) Karim Mazaheri

More information

Kevin J. Barker. Scott Pakin and Darren J. Kerbyson

Kevin J. Barker. Scott Pakin and Darren J. Kerbyson Experiences in Performance Modeling: The Krak Hydrodynamics Application Kevin J. Barker Scott Pakin and Darren J. Kerbyson Performance and Architecture Laboratory (PAL) http://www.c3.lanl.gov/pal/ Computer,

More information

ARCHITECTURE SPECIFIC COMMUNICATION OPTIMIZATIONS FOR STRUCTURED ADAPTIVE MESH-REFINEMENT APPLICATIONS

ARCHITECTURE SPECIFIC COMMUNICATION OPTIMIZATIONS FOR STRUCTURED ADAPTIVE MESH-REFINEMENT APPLICATIONS ARCHITECTURE SPECIFIC COMMUNICATION OPTIMIZATIONS FOR STRUCTURED ADAPTIVE MESH-REFINEMENT APPLICATIONS BY TAHER SAIF A thesis submitted to the Graduate School New Brunswick Rutgers, The State University

More information

Optimization of Laminar Wings for Pro-Green Aircrafts

Optimization of Laminar Wings for Pro-Green Aircrafts Optimization of Laminar Wings for Pro-Green Aircrafts André Rafael Ferreira Matos Abstract This work falls within the scope of aerodynamic design of pro-green aircraft, where the use of wings with higher

More information

This is an author-deposited version published in: Eprints ID: 4362

This is an author-deposited version published in:   Eprints ID: 4362 This is an author-deposited version published in: http://oatao.univ-toulouse.fr/ Eprints ID: 4362 To cite this document: CHIKHAOUI Oussama, GRESSIER Jérémie, GRONDIN Gilles. Assessment of the Spectral

More information