Inhaltsverzeichnis
Exact Diagonalization of Large Sparse Matrices: A Challenge
for Modern Supercomputers
Outline
Physical Motivation: Strongly correlated charge-lattice models
Physical Motivation: From the experiment to supercomputers
Numerics: Exact diagonalization
Numerics: Sparse MVM techniques
Numerics: Numerical core of JDS format
Numerics: Scalability of the implementation
Performance study: Benchmark platforms
Performance study: Single processor characteristics: RISC
Performance study: Single processor characteristics: Vector/SMP
Performance study: CRAY T3E vs. Origin3800
Performance study: CRAY T3E vs. Vector/SMP
Summary
Acknowledgement
|
Autor: Gerhard Wellein
E-Mail: gerhard.wellein@rrze.uni-erlangen.de |