Parallel implementation of block boundary value methods for ODEs (Q676146)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Parallel implementation of block boundary value methods for ODEs
scientific article

    Statements

    Parallel implementation of block boundary value methods for ODEs (English)
    0 references
    0 references
    0 references
    18 August 1997
    0 references
    The initial value problem is transformed to a block boundary value problem (BVP) over the domain of solution, using two \(k\)-step linear multistep formulas on a coarse and fine grid. This means simultaneous (instead of recursive) solution over \(t\). However, the ordinary differential equation (ODE) must be linear. The resulting solution of a linear system is parallelized. The method can also be used for BVPs. Excellent speedup is obtained on a transputer-based distributed memory computer. I miss the comparison to the recursive method.
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    parallel implementation
    0 references
    block boundary value methods
    0 references
    parallelization
    0 references
    linear multistep formulas
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references