High-level synthesis: technology transfer to industry

During the last decade, several university projects have addressed the problem of synthesizing the architecture of digital integrated circuits from descriptions of their behavior. The authors explain how they extended one research project, the System Architect's Workbench (SAW) from Carnegie-Mellon University, into a tool that designers can use. Significant enhancements were made to the SAW to enable it to make architectural tradeoffs and simulate the designs it produces. The resulting architecture design tool was tested on a Texas Instruments' design. Based on their experience, the authors discuss how close high-level synthesis research is to being industrially useful.<
>