During the last decade, several university projects have addressednthe problem of synthesizing the architecture of digital integratedncircuits from descriptions of their behavior. The authors explain hownthey extended one research project, the System Architect's Workbenchn(SAW) from Carnegie-Mellon University, into a tool that designers cannuse. Significant enhancements were made to the SAW to enable it to makenarchitectural tradeoffs and simulate the designs it produces. Thenresulting architecture design tool was tested on a Texas Instruments'ndesign. Based on their experience, the authors discuss how closenhigh-level synthesis research is to being industrially useful
展开▼