Architectures, components, and software tools for system-on-chip design and implementation

点击翻译按钮获取中文摘要

Improvements in microelectronics fabrication technology continue to enable the integration of many components onto a single chip for use in a variety of applications such as computer systems, portable telephones and other consumer products, and sophisticated biomedical devices. This type of integration has many benefits including higher performance, reduced power consumption for longer battery life in portable and embedded systems, and physically smaller devices and systems. In order to aid the designers of single-chip systems in addressing the complexity of the task of developing such systems, the proposed research will investigate novel architectures, components, and software tools for implementing highly-integrated chips. Architectures define how the processing, memory, and other components are logically organized for implementation within a single chip. The components themselves can be customized to provide the appropriate functionality to suit a particular application. Finally, sophisticated software tools can aid designers in selecting, configuring, and combining the appropriate components into the most suitable architectures for implementation. Given the high complexity of single-chip systems, the aim is to automate as much of the entire process from architectural conception to detailed implementation as possible. Because the actual implementation of highly-integrated custom chips to characterize the benefits of new architectures, components, and software tools is an expensive undertaking, the proposed research will rely on the use of prefabricated, programmable chips and the associated software tools for converting design specifications into hardware configurations for these chips. Advances in microelectronics fabrication technology have given these programmable chips the same benefits as customized chips, hence the growing capacities of programmable chips are sufficient for prototype implementation of large single-chip designs that will be generated in the pursuit of this research. The experience and results that are obtained from this approach can then be applied to the implementation of custom chips.