Fine-Grain MPI Year Two
The demand for computing power has been rapidly growing over the last decade. The ability to efficiently utilize computing resources and improve the productivity of applications is necessary for the competitiveness of any industry and it will become more critical as the demand for computational resources increases. The computer hardware sector has seen rapid advances with the introduction of multicore and many-core processors and has introduced many challenges for the software development community to efficiently utilize the new architectures. My research is focused on developing systems and techniques that allow improved utilization of parallel machines. The Message Passing Interface (MPI) is extensively used for constructing parallel applications in distributed environments. Our work is related to development of systems for high performance computing through extending and enhancing the capabilities of the MPI middleware. We have developed a system called Fine-Grain MPI that allows execution of hundreds and thousands of MPI processes on-chip or communicating between chips inside a cluster. The potential benefits of our approach are for multicore as the number of cores increase.