Speculative supplier identification for reducing power of interconnects in snoopy cache coherence protocols

摘要

It is our pleasure to introduce the 2007 edition of the International Conference on Computing Frontiers. The goal of this conference is to peer into the future of computing, to expose researchers in various computing disciplines to each others' work, and to explore extreme and novel ideas pushing computing to its limits. This year's program spans a wide range of topics, including technology-driven, embedded and highperformance architectures, massively parallel systems, software for emerging systems, and quantum computing. Besides the regular sessions, we have added three keynotes on topics that reflect the contents of the proceeding. >In order to have a reliable evaluation of research papers in such a wide variety of research areas, we decided early on to organize the program committee in six tracks, each headed by a world-class expert acting as vice program chair. We were very fortunate to be able to convince the best people our community can offer to act as vice chair. The vice chairs recruited program committee members specialized in the area covered by their track. As you can tell, the list of program committee members is long and distinguished. By delegating the decision process to vice chairs we were able to get recommendations from experts in each area of research. >After the paper submission deadline, we assigned each paper to the most appropriate track. To provide a substantive and fair evaluation of each paper, each paper was assigned to at least four program committee members and, in some cases, a paper was also reviewed by external reviewers. At the end we collected an average of 4.3 reviews per paper (242 reviews for 56 papers). After recommendations from the vice chairs and deliberations across all tracks, we decided to accept 28 papers.