Welcome to My Personal Website!
Current Research Topics
- High-Performance Computing using GPUs
- Scientific Computing on GPUs
- Wireless Networking
My current research is high-performance computing using graphics processors. We are currently working on applications in both Chemical Engineering and Physics. You can find more information about our work on Monte Carlo Simulation using GPUs at our GOMC web site. By developing novel algorithms that exploit the massively multi-threaded architecture of the GPU as well as co-exist within the hierarchical memory model of modern GPUs, we are able to significantly reduce the running time of such applications. Our primary goal is to leverage the GPU to obtain significant speedups for scientific codes. We are also interested in getting a better understanding of how to best optimize performance on GPUs and how to automate this optimization process. To conduct research with me in this area, students need to have solid skills in algorithm development and software implementation. A strong foundation in mathematics is especially helpful.
My previous research on wireless sensor networks and embedded systems addressed problems related primarily to networking applications such as vehicular networks and biomedical (body-area) networks. The research in our lab also considered more general applications of wireless sensor networks, such as routing, self-organization, and resource utilization. I currently have one or two Ph.D. students completing their dissertations in this area. For personal and professional reasons, I have decided not to continue working in this area and instead focus my research on high-performance computing and other research areas. Therefore, I'm not accepting new students with research interests in this area.
In addition to the scientific applications mentioned above, we are also interested in general-purpose computing on GPUs. Our objective is to apply our knowledge of efficient programming of GPUs to a wide variety of computational problems that can benefit from the high degree of cost-effective parallelism offered by the GPU. I also still continue some efforts on the use of wireless networking for medical applications. This is joint research with, for instance, the College of Nursing. Although this is an active area of research, I'm not recruiting Ph.D. students in this area.