Home

--- Think Parallel ---

This project aims at exploiting the massive parallel processing capability of the rapidly emerging multi-core processor systems. There is broad consensus within the microelectronics industry that processors with 100s of cores will be available within the next 5 to 10 years, providing supercomputing capabilities on a single chip, and allowing us to solve much larger instances of computationally intractable problems from nearly every scientific and engineering discipline. However, there is a huge gap between the potential performance that can be delivered by multi-core systems and our ability to program them to actually realize the performance gains. This proposal is aimed at developing a comprehensive software infrastructure that will bridge this ever widening gap.