Miss penalty reduction using bundled capacity prefetching in multiprocessors | IEEE Conference Publication | IEEE Xplore