 |
|
|
|
|
|
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DARPA | Scalable Systems and Software | Budget Code: | ST-19 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
The Scalable Systems and Software program supports the development of computing and advanced software technologies needed to enable the development, introduction, and effective use of secure scalable and distributed high performance computing technologies. The program accelerates the evolution of computing and communications technologies. The high performance computing technologies developed by this program drive national defense capabilities in areas that benefit from the ability to do complex modeling and secure distributed communication.
Prototype Scalable Systems is focused on developing, demonstrating, and transferring the scalable computing systems technology base. Projects in this area cover a broad range of computing systems technology issues from research into future generation computing systems technology to the validation of scalable concepts through supporting the incorporation of advanced technology into industrial prototypes. Operating Systems and Services supports the development of advanced software technologies to enable the development, introduction, and effective use of secure, scalable and distributed high performance computing technologies. In the past, it was sufficient to develop a "fast" computer system in the absence of a software environment. This is no longer the case as the time between generations of advanced hardware has been reduced from many years to less than two years, making it essential to maintain a stable software development environment that can span several generations of computer hardware. This is an absolutely mandatory need for DoD which cannot afford to redevelop software for each new generation of computers. |
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Milestone Changes | FY 96 milestone for scalable OS moved from System Environments | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| FY 1995 Actual Milestones | FY 1996 Estimated Milestones | FY 1997 Agency Requested Milestones | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Designed system architectures incorporating components such as programmable protocol engines to support scalability and high performance.
Demonstrated systems tools for on-line analysis of a real-time operating systems for scalable, distributed HPC systems. Demonstrated operating system ability to confine processes to isolated domains. Demonstrated first HPC single node operating at 1 GFLOP/s. |
Demonstrate high-availability systems scalable in performance to 1 teraFLOP/s.
Demonstrate extensible modular operating system framework supporting real-time, distributed, and limited fault-tolerant scalable computing applications. Demonstrate user-extensible microkernel operating system technology, integrating compiler and run-time support services. Demonstrate computing node architectures that dramatically increase internal memory and communications bandwidths. Demonstrate I/O enhancements to a scalable operating system that overcomes identified bottlenecks leading to significant improvements in throughput. Evaluate first generation of fully scalable OS software and programming environments on small-scale versions of teraops computing systems. |
Demonstrate scalability from distributed workstation clusters to teraFLOP supercomputers on the identical technology base.
Demonstrate distributed cluster technology scalable to teraFLOPS. Demonstrate advanced object management systems integrated with operating systems and applications to achieve efficient use of memory while enhancing execution speed. Demonstrate the prototype of a scalable operating system that incorporates high assurance capabilities for the Defensive Information Warfare program Demonstrate extensible modular operating system framework supporting real-time, distributed, and limited fault-tolerant scalable computing applications. |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |