Georgia Tech Embedded Systems Lab

An overview of our work

We believe that it is important to work in architecture, methodology, and applications in order to conduct cutting-edge research in embedded systems.

We have long used video as a prime application area for our embedded systems research. Video requires very high computation rates and large amounts of memory; video systems are often cost-sensitive as well. Because video systems are built as systems-on-chips, they make a good arena for investigating both architectures and methodologies for embedded systems.We're also beginning to look into radio and communications as a fertile application area for embedded computing.

We have worked on tools and methodology for quite some time. This work started out as hardware/software co-design in the strict sense, meaning synthesizing multiprocessors and task-level software for embedded systems. More recently, we've expanded our work into methodologies for platform-based design.

We work on two types of architecture. One is large-scale application-specific architectures for video. We have worked on video signal processors (VSPs) for quite some time. We have also worked on accelerators for video operations. Our original work was motivated by MPEG-style video compression; we are now trying to extend that work to architectures for smart cameras. The other type of architecture we work on relates much more closely to logic design. Code compression and bus encoding are examples of this area of research. Code compression and bus encoding directly affect computer architectures, but they also rely on a sound knowledge of logic design and encoding algorithms.

Projects

For more details on our research, please see the projects page.