Massively Parallel Systems Design for Real-Time Embedded Applications
Abstract
This paper describes a generic approach to mitigate risk when reengineering for high throughput massively parallel systems. The approach entails baselining the existing system, capturing the functional requirements, estimating initial processing requirements through a high level analysis, benchmarking a subset of the functionality on a low throughput computer, and modeling the high throughput application to determine the detailed processing requirements for scaling.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 14, 1994
- Accession Number
- ADA277256
Entities
People
- Chin-hwa Lee
- Thomas C. Choinski
Organizations
- Naval Undersea Warfare Center