Modeling cache performance beyond LRU | IEEE Conference Publication | IEEE Xplore