 |
|
|
|
|
|
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DOE | High Performance Computer Systems | Budget Code: | KC 07 01 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
This activity funds research into advanced computer architectures and novel commercial prototypes suitable for scientific computing. Funding for development of new prototype architectures has been deemphasized in favor of providing feedback to commercial vendors of novel systems. This activity does include funding to develop a message passing interface for an MPP system, which is critical for easily porting codes among massively parallel computers.
This program activity also helps fund the acquisition of large prototype computers for testing and evaluation. Grand Challenge-class problems require large prototype systems because they cannot be scaled down without removing essential aspects of their physics. |
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Milestone Changes | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| FY 1995 Actual Milestones | FY 1996 Estimated Milestones | FY 1997 Agency Requested Milestones | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Produced the reference implementation for a message-passing interface on massively parallel processors (deliverable from IBM-ANL collaboration using SP1).
A joint agency project (with NASA, NSF, and ARPA) was initiated in scalable I/O for parallel computers. Installed Intel XP/S 150 Paragon (1024 nodes, 3072 processors) at ORNL. Had full-scale CM-5 and Paragon systems operational and supporting Grand Challenge research projects. Started workshops and projects to investigate next-generation computational technology for architectures, computational environments, and applications to achieve petaflops performance required by DOE programs such as Science Based Stockpile Stewardship. |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |