The primary focus of my work is flash storage. This is the first new storage technology in most of our lifetimes which offers serious competition to magnetic disk, and the introduction of low-latency solid-state storage creates opportunities for entirely new ways of integrating storage and computation. The three main thrusts of my research in this area are:
Flash-based Active Storage. The characteristics of Flash (and future solid-state technologies) allow us to create parallel arrays of processors and storage in ways never possible with disk. Is it possible to harness the power of such an architecture in a way that goes beyond niche applications and is useful for general-purpose problems?
Hybrid Flash+DRAM main memory. By combining Flash and a small amount of cache, non-volatile devices have been proposed which provide DRAM interfaces, with lower speed but much higher density and lower power. Using such memory requires radically different memory management and storage architectures than are found in today's operating systems.
Object persistence systems for low-latency and flash storage.
If you are a talented student who is interested in any of these areas, please send me an email to see if there are any opportunities to work together.