This allows us to overlap bringing the block into cache with the search for the next available block.