G
Graham F. Carey
Researcher at University of Texas at Austin
Publications - 253
Citations - 6032
Graham F. Carey is an academic researcher from University of Texas at Austin. The author has contributed to research in topics: Finite element method & Mixed finite element method. The author has an hindex of 37, co-authored 253 publications receiving 5803 citations. Previous affiliations of Graham F. Carey include University of Texas System.
Papers
More filters
Proceedings ArticleDOI
A scalable, object-oriented finite element solver for partial differential equations on multicomputers
TL;DR: A parallel finite element solver for mesh-connected multicomputers (ie., message-passing, distributed-memory multiprocessors) is presented and a theoretical proof that the algorithms used are highly scalable and close to optimal is provided.
Journal ArticleDOI
An enhanced polygonal finite-volume method for unstructured hybrid meshes
TL;DR: In this paper, a polygonal control volume is proposed for irregular hybrid meshes for incompressible viscous flow past a circular cylinder, which is shown to be superior to the conventional node-dual finite-volume discretization.
Journal ArticleDOI
Preconditioners for high degree elements
E. Barragy,Graham F. Carey +1 more
TL;DR: The present investigation considers element block preconditioners for discretizations employing elements of high degree for regular grids and for grids with non-constant coefficients or highly skewed grids, the block precONDitioners described here can be applied instead of Jacobi preconditionsing.
Journal ArticleDOI
A distributed memory parallel element-by-element scheme for semiconductor device simulation
S.W Bova,Graham F. Carey +1 more
TL;DR: A domain decomposition and parallel element-by-element (EBE) scheme is developed for semiconductor device simulation modeled by the drift-diffusion (DD) equations and a classical Gummel iterative decoupling of the potential and carrier transport equations is applied.
Journal ArticleDOI
Maximizing Sparse Matrix-Vector Product Performance on RISC Based MIMD Computers
TL;DR: This work shows that standard approaches with Fortran or C may not deliver good performance and presents a strategy involving managing the cache to improve the performance and demonstrates that it is possible to achieve 2.5 times better performance over a Fortran implementation with an assembly coded kernel on an Intel i860.