The multicore revolution is underway. Classical algorithms have to be revisited in order to take hierarchical memory layout into account. In this paper, we aim at minimizing the number of cache misses paid during the ...