On the parallelization of an adaptive multigrid algorithm for a class of flow problems

Conference · Tue Jan 01 04:00:00 EST 1991

OSTI ID:5912043

Thompson, C P ^[1]; Cowell, W R; Leaf, G K ^[2]

---

1. Bergen Scientific Centre (Norway)
2. Argonne National Lab., IL (USA)

The authors describe the parallelization of an adaptive multigrid code that solves the steady, laminar, incompressible Navier-Stokes equations. Parallelism was achieved as loop concurrency on a shared-memory multiprocessor. Analysis of the code for parallelism, in particular the location of data dependencies that inhibit parallelism, was aided by the use of the software tool VecPar-77. They propose a model of loop-based parallel computation that enables them to define the relationship among the notions of speed-up, extent of parallelization, and problem size. They present speedup results obtained on a Sequent Symmetry and use the model to interpret them. 4 refs., 6 figs., 1 tab.

• Research Organization: Argonne National Lab., IL (USA)
• Sponsoring Organization: DOE; USDOE, Washington, DC (USA)
• DOE Contract Number: W-31109-ENG-38
• OSTI ID: 5912043
• Report Number(s): ANL/CP-72220; CONF-9103161--1; ON: DE91010795
• Country of Publication: United States
• Language: English

Related Subjects

640410* -- Fluid Physics-- General Fluid Dynamics
75 CONDENSED MATTER PHYSICS
SUPERCONDUCTIVITY AND SUPERFLUIDITY
99 GENERAL AND MISCELLANEOUS
990200 -- Mathematics & Computers
ALGORITHMS
ARRAY PROCESSORS
COMPUTER ARCHITECTURE
COMPUTER CODES
DIFFERENTIAL EQUATIONS
EQUATIONS
FLUID FLOW
FLUID MECHANICS
INCOMPRESSIBLE FLOW
LAMINAR FLOW
MATHEMATICAL LOGIC
MATHEMATICAL MODELS
MECHANICS
MESH GENERATION
NAVIER-STOKES EQUATIONS
PARALLEL PROCESSING
PARTIAL DIFFERENTIAL EQUATIONS
PERFORMANCE
PROGRAMMING
STEADY FLOW