Expo/Science & Industry/Spacetime Wrinkles
| Back 1 |
Back 2 | Map | Glossary | Information |
Spacetime Spinoffs
As numerical relativity researchers evolve new computational methods, the tools and techniques they devise should prove valuble in seemingly
unrelated areas of science and technology.
Though the Einstein equations are very complex, they share the following common features with important equations embedded in other science and engineering problems:
- The equations are very general hyperbolic and elliptical systems.
- They deal with wave propagation.
For example, some of the algorithms that rapidly solve the equations to high accuracy on parallel machines could also find applications in the oil industry. Industry seismologists need to understand how sound waves propagate through the Earth's crust.
Tackling this Grand Challenge requires that relativity scientists work closely with computer scientists and computer manufacturers to advance the state-of-the-art in the following areas:
- Efficient numerical algorithms for parallel computers
- New methods for solving hyperbolic equations
- New methods for solving the mathematics of giant linear systems and elliptic equations
- Improved techniques for solving constrained systems
- Better methods for writing and processing huge (several gigabyte) files produced on parallel computers.
- For supercomputers, more efficient compilers, the software needed to translate computer code into machine-executable operations
- More advanced ways of visualizing complex scientific data, including
- Virtual reality
- Superior approaches to heterogeneous computing, the linking of diverse types of networked
computers at different locations in order to solve problems in an optimal fashion.
Numerical relativity is particularly demanding in every one of these areas. Developing techniques to solve Einstein's equations will have broad impact throughout computational science and engineering.
Progress in numerical relativity will benefit other areas of astrophysics. Ultimately, researchers will develop codes
that solve the complete, coupled spacetime-fluid equations that govern the evolution of all astrophysical systems, such as
supernovae,
neutron stars, accretion disks,
quasars, and the universe itself.
Return to 3D and Beyond
Return to Code Building
Exhibit Map
Glossary
Information Center
Copyright © 1995, The Board of Trustees of the University of
Illinois
Ed Seidel, David Curtis, NCSA. Last modified 9/22/95.