ISCA

It is not necessary to extol the importance of ISCA to the computer architecture research community. ISCA has been for some time our flagship symposium, and there is no reason to believe this won't continue to be the case. This year, Dirk Grunwald and his program committee culled 27 papers from an almost record number of submissions to put together the technical program. I hope you find some of these papers thought-provoking, inspiring, and useful to your needs.We have continued the ISCA tradition, started in 1989 in Eilat by Doug DeGroot, of holding workshops (along with tutorials) the weekend before the symposium. Sanjay Patel, the Workshops/Tutorials Chair, has put together a collection of four workshops and four tutorials on particularly relevant topics. We have changed the registration procedure associated with workshops and tutorials this year to make it easier for attendees to sample all that is offered. I hope you can take advantage of them as well.I am particularly pleased to welcome you to Anchorage. When we first suggested Alaska as the site for ISCA, some were skeptical. We politely listened to all the reasons why Alaska wouldn't work, and with the support of the many, we pushed forward. The result, I think you will find, is that our flagship symposium is being hosted in a flagship location. We will attempt to give you a taste of the majesty of Alaska during the conference, but we do not really have enough time to do it properly. I hope you are able to stay the weekend and continue where we leave off.We received a total of 180 submissions for the conference, of which 27 were selected for the final program. The low acceptance rate (15%) reflects the growing number of submissions to ISCA, since the number of accepted papers was not significantly different than in the past. Each paper was reviewed by at least four reviewers, and at least two of those reviews were by program committee members. A total of 300 people helped review papers. All reviewing was double-blind. The review process was slightly different than in the past; authors were allowed to see and respond to reviews prior to the program committee meeting.The actual program committee meetingwas on January 25th, but the committee had already done a great deal of work. The author responses were made available to the program committee and despite the huge volume of additional work, the committee and program chair examined each rebuttal. Furthermore, papers that had any issues concerning reviews raised by authors were specifically discussed at the program committee meeting. Papers that involved program-committee member conflicts were handled using a "hot seat" method where the conflicting member was asked to leave the room. Equally important, authors were given the option to let reviewers see the outcome, other reviews and author responses for their papers. Our goal was to help authors educate reviewers; some reviewers had amazingly detailed reviews while others provided only cursory feedback.