2016CS 267 Applications of Parallel Computers.pptVIP

CS 267 Applications of Parallel ComputersSolving Linear Systems arising from PDEs - I James Demmel /~demmel/CS267_2002_Poisson_1 Outline Review Poisson equation Overview of Methods for Poisson Equation Jacobi’s method Red-Black SOR method Conjugate Gradients FFT Multigrid (next lecture) Recap of “Sources of Parallelism” Lecture Discrete event systems: Examples: “Game of Life,” logic level circuit simulation. Particle systems: Examples: billiard balls, semiconductor device simulation, galaxies. Lumped variables depending on continuous parameters: ODEs, e.g., circuit simulation (Spice), structural mechanics, chemical kinetics. Continuous variables depending on continuous parameters: PDEs, e.g., heat, elasticity, electrostatics. A given phenomenon can be modeled at multiple levels. Many simulations combine more than one of these techniques. Continuation of Recap Examples of PDE systems include Hyperbolic problems (waves): Quantum mechanics: Wave-function(position,time) Elliptic (steady state) problems: Electrostatic or Gravitational Potential: Potential(position) Parabolic (time-dependent) problems: Heat flow: Temperature(position, time) Diffusion: Concentration(position, time) Many problems combine features of above Fluid flow: Velocity,Pressure,Density(position,time) Elasticity: Stress,Strain(position,time) Terminology Term hyperbolic, parabolic, elliptic, come from special cases of the general form of a second order linear PDE a * ?2u/?x2 + b * ?2u/?x?y + c * ?2u/?y2 + d * ?u/?x + e * ?u/?y + f = 0 (y denotes time for hyperbolic and parabolic equations) Analog to solutions of general quadratic equation a * x2 + b * x*y + c * y2 + d * x + e * y + f = 0 Ellipse: 4*a*c – b^2 0 Hyperbola: 4*a*c – b^2 0 Parabola: 4*a*c – b^2 = 0 Solving PDEs Solution strategies: Hyperbolic problems (waves): Use explicit time-stepping Solution at each point depends on neighbors at previous time Elliptic (steady s