 |
|
|
|
|
|
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| NASA | Grand Challenge Support | Budget Code: | 509-10,-20,-30,505,535-538,212,232,233,306,656 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
This research area develops, enhances, and evaluates techniques for the multidisciplinary modeling and simulation required by Grand Challenge problems. The Computational Aeroscience (CAS) research will focus on understanding the high performance computing environment and how it can be used to solve a range of problems of importance in aerospace engineering at a cost that represents the value, flexibility, and short cycle time required by the aerospace community. This will be accomplished with the aid and focus provided by several representative aerospace design problems related to nationally important programs. These are: a High Speed Civil Transport (HSCT); an Advanced Subsonics Civil Transport (ASCT); High-Performance Aircraft (HPA); and Rotorcraft.
Earth and Space Science (ESS) research will cover two critical scientific areas: the coupling of advanced discipline models into scalable global simulations providing realistic global change understanding; and the integration of models and analysis algorithms for processing, analyzing and understanding the enormous volumes of data expected from scientific missions. Research will focus on: large scale structure and galaxy formation; cosmology and accretion astrophysics; convective turbulence and mixing in astrophysics; solar activity and heliospheric dynamics; Earth system models; four-dimensional data assimilation; climate models, and knowledge discovery in geophysical databases and satellite data. Collaborative groups including discipline scientists, software and systems engineers, professional software developers and algorithm designers are supported by shared computational and experimental facilities. Technical accomplishments will include development of application-specific codes for innovative high- performance computing systems, design and analysis of algorithms, and architecture and performance assessment of specific applications. NASA will evaluate the early research products and make the results available to system vendors as quickly as possible. Results in design and theory of algorithms are as important to breaking down computational scaling barriers as are performance improvements in computing hardware. Algorithms for common techniques such as multidimensional FFTs, Fast Poisson solvers, multigrid methods, Reimann solvers, sparse matrix methods, singular value decomposition, matrix factorization methods, and spectral methods are being re-implemented on a variety of architectures in order to understand how architecture affects efficiency and algorithm design. |
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Milestone Changes | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| FY 1995 Actual Milestones | FY 1996 Estimated Milestones | FY 1997 Agency Requested Milestones | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Developed multidisciplinary Earth and space science applications on 10-50 GigaFLOPS testbed, including: Cosmology and accretion astrophysics, large scale structure and galaxy formation, convective turbulence and mixing in astrophysics, solar activity and heliospheric dynamics, atmosphere/ocean dynamics and tracers chemistry.
The Earth and Space Science Team evaluated testbed architectures, running their scientific problems, for usability and efficiency. Developed advanced subsonic civil transport analysis/optimization code and perform medium fidelity high speed civil transport modeling. Enhanced the ability of aircraft manufacturers to quickly analyze different design options and accelerate the prototyping process, thereby reducing design cycle costs and producing vehicles with improved performance. Demonstrated 32 GigaFLOPS sustained performance on 144 nodes of the IBM SP-2 at NASA Ames in support of FY96 milestone to demonstrate multidisciplinary aeroscience applications on 10-50 GigaFLOPS testbeds. |
Demonstrate multidisciplinary aeroscience and Earth and Space Science applications on 10-50 GigaFLOPS testbeds.
Demonstrate end-to-end reductions in cost and time to solution for aerospace design applications on heterogeneous systems. The approach was to make a CAS testbed of heterogeneous workstations and high speed local networking openly accessible to software activities similar to existing HPCC work on TeraFLOPS systems. Design applications, oriented around the propulsion industry needs, guided the system software. Select and award Round-2 ESS Investigators. |
Demonstrate multidisciplinary Earth and Space Science applications sustaining 50-100 GigaFLOPS.
Complete TeraFLOPS scalable ESS and CAS Parallel Benchmarks. |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |