Michael Spevak

Learn More
A generic scientific simulation environment is presented which imposes minimal restriction regarding topological, dimensional, and functional issues. Therewith complete discretization schemes based on finite volumes or finite elements can be expressed directly in C++. This work presents our multi-paradigm approach, our generic libraries, some applications(More)
A three-dimensional unstructured mesh adaptation technique coupled to a posteriori error estimation techniques is presented. In contrast to other work [1,2] the adaptation in three dimensions is demonstrated using advanced unstructured meshing techniques to realize automatic adaptation. The applicability and usability of this complete automation are(More)
We discuss discretization schemes for the Poisson equation, the isothermal drift-diffusion equations, and higher order moment equations derived from the Boltzmann transport equation for general coordinate systems. We briefly summarize the method of dimension reduction when the problem does not depend on one coordinate. Discretization schemes for(More)
We present an orthogonal topological framework which is able to provide incidence traversal operations for all topological elements. The run-time performance of this topological traversal operations can be optimized at a highly expressive level, where the abstraction penalty imposed by this approach is negligible. For the topological storage we use(More)
This work analyzes the performance of high-precision interconnect simulation tools on refined meshes with guaranted accuracy. On the one hand, the integrated circuits are subject to an ongoing miniaturization which results in ever increasing computing power. On the other hand, the simulation of these integrated circuits demands more sophisticated simulation(More)
We present a novel error estimation driven threedimensional unstructured mesh adaptation technique based on a posteriori error estimation techniques with upper and lower error bounds. In contrast to other work [1, 2] we present this approach in three dimensions using unstructured meshing techniques to potentiate an automatically adaptation of(More)
We present a high performance environment for scientific simulation applications (GSSE). This environment is based on the three orthogonal concepts of topology, geometry, and quantities. Lambda calculus is used in order to assemble various partial differential equations for TCAD and other physical equations. We present examples from device and process(More)
We present an error estimation driven three-dimensional unstructured mesh optimization technique based on a posteriori error estimation techniques. We briefly summarize error estimation techniques as well as mesh quality estimations and apply them to problems of semiconductor device and process simulation. We then optimize the tessellation of the simulation(More)