Clusters built from commodity hardware and software have several advantages over more traditional supercomputers. Commodity clusters are cheap and ubiquitous, and so they are easier to design, program and maintain. However, as high-speed networks reach 10Gb/s speeds and modern computer architectures reach past 1GHz, new commodity clusters are unable to harness this power.
Our research goal is to explore how we can splinter commodity operating system resources in order to maximize their performance. Splintering is the process of determining which functionality to extract from the kernel core on a host processor and determining to where in a system the functionality should be moved to improve performance for a given application and architecture.