|
|
Projects
|
| Search - content lookup - is a key operation in many real-world applications, such as a high-performance network router and a large-scale speech recognition system. In this research, we aim to design a search-capable memory architecture that is high performance, low power, and low cost. |
| Multicore processors (having multiple processor "cores" on a single chip) are everywhere, and we will find chips having many more cores in the future. A key question for the next-generation "many-core" processor is: How can we utilize the increased amount of on-chip resources (cores, caches and interconnects) to achieve scalable performance, low power, high availability and dependability? We aim to provide a flexible many-core processor architecture in this research to answer the question. |
| Various reliability issues have manifested; for example, increased process variation may cause unexpected circuit failures. Advanced processor chip yields are staggering due to aggressive scaling and high complexity involved in design and manufacturing process. In this research, we are investigating how design practices have to be changed and how processors need be "reconfigured" to overcome serious reliability and yield erosion problems. This project is supported by a grant from the NSF. |
| In this research, we tackle application-specific design problems in various embedded system domains. Currently, issues around new hybrid hard drive (HHD) and coarse-grained reconfigurable architecture (CGRA) are investigated. We are particularly interested in reducing power consumption. |
Acknowledgement: We are using simics and simplescalar simulation tools for our research study.