Issue No. 01 - January/February (2007 vol. 27)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/MM.2007.23
David H. Albonesi , Cornell University
Welcome to the first issue of IEEE Micro for 2007! It's an honor and a privilege to serve the Micro community as the new Editor in Chief. When I first subscribed to Micro—as a design engineer in the computer industry, perhaps 20 years ago—I remember thumbing through the current issue as soon as it arrived in the mail to see which articles most piqued my interest. Fast-forward to today, and things are pretty much the same: I still subscribe to the paper edition, and much to the dismay of my wife and kids, I eagerly grab it out of the mail pile when it arrives! (Yes, I know, I need to get a life.)
So when the opportunity arose to be considered successor to Pradip Bose as Micro EIC, I jumped at the thought of leading what I have long considered the best periodical combining leading microprocessor research with real industry designs. My enthusiasm abated when the reality of attempting to fill Pradip's rather large shoes set in. I've known Pradip for about 10 years now, and he has been very supportive of my career and those of many others in our field. He is truly a giant in the microarchitecture community (although in his quiet humility he would scoff at such an assertion). He has coauthored many highly cited papers (in recent years, seminal papers in power- and reliability-aware microarchitecture) and has tirelessly served the computer architecture community in many capacities, despite a hectic schedule as a manager at IBM Watson. I realized that Pradip's inspirational leadership as EIC of Micro would be very difficult to match given his many accomplishments. A few examples: He instituted the highly popular Top Picks issue; he infused new blood into the Editorial Board; he sponsored IEEE Micro Best Paper Awards at top conferences; he established an Advisory Board comprising industry and academic leaders in the microarchitecture community; he engaged both boards in setting a more focused direction for Micro; and he worked behind the scenes to rightfully defend Micro as a highly relevant magazine for a vitally important IEEE Computer Society constituency. The entire Micro community should flood Pradip's inbox (many have already done so) with thanks for all he has done for this vital IEEE CS periodical. (Thankfully, we can continue to count on Pradip's leadership, owing to his gracious acceptance of my invitation to join the Micro Advisory Board.)
Ultimately, I realized that given all that Pradip accomplished as Micro EIC, I could not ask for more solid ground to stand on in undertaking the challenge of my first EIC position. Indeed, one of the greatest advantages of following Pradip as EIC is that much of the planning for 2007 is already done! We are currently putting the finishing touches on the March/April Hot Chips issue, which, as in past years, will contain full-length technical articles from presentations made at the Hot Chips conference. The May/June theme is Hot Tutorials-another brainchild of Pradip's, based on input from the Advisory Board; the issue will feature tutorial-style articles covering a variety of timely topics. Following a general-interest issue in July/August, we are planning an exciting September/October special issue on Interconnects for Multicore Chips, coedited by Li-Shiuan Peh of Princeton and Partha Kundu of Intel. I encourage academic and industry groups working in this critical research area to read the Call for Papers in this issue and to submit an article on their work.
With Pradip's many accomplishments to build on, outstanding Advisory and Editorial boards in place, a dedicated Micro staff, and no shortage of interesting microarchitecture research and microprocessor chip designs to showcase, Micro is well positioned for the future. I am personally very bullish on the microarchitecture landscape that lies ahead of our community. The field is ripe for innovation as Moore's Law continues to deliver new opportunities, yet significant challenges. The strong industry push towards "many-core" microprocessors with tens, and perhaps hundreds, of processor cores on a die has led to a resurgence in multiprocessor systems research (but with the significant difference that the cores and much of the memory hierarchy are integrated in a single 3D package); there is a dire need to address issues such as programmability and debugging in these future chips. The question of how we create systems-on-a-stack that can efficiently exploit parallelism in a wide variety of software, yet with manageable design and verification complexity, is one that the Micro community will be vigorously debating in the next few years. Part of this debate will center on special-purpose chip functionality to accelerate human-computer interface (HCI) functionality, security, physical simulation, and other contemporary applications. As the decades-long practice of worst-case circuit design becomes increasingly unviable with continued technology scaling, we will need complexity-effective runtime microarchitectural techniques that address power/thermal, reliability, and extreme static and dynamic variations. These are but a few areas of interest—I'm sure you have a number of others on your own personal lists.
The current issue in fact highlights some of the best research emerging from our academic and industry laboratories in multicore design, and in other areas of vital importance to the Micro community. The Top Picks issue has evolved from an undertaking of a few Editorial Board members to one of a full-fledged program committee today. We were extremely fortunate to have Ronny Ronen of Intel and Antonio González of Intel and UPC team up to lead this year's effort. As a member of their program committee, I witnessed first-hand their painstaking efforts as co-chairs. As one might imagine, picking 11 papers from a strong batch of 76 submissions (all previously selected for top computer architecture conferences) is exceedingly challenging—to do it as efficiently and fairly as possible requires tremendous effort and dedication. Ronny's and Antonio's tireless efforts resulted in 100 percent return of requested reviews and 100 percent attendance at the program committee meeting. The entire evaluation process was as well run as any that I have been involved in, and I think you will find that the papers in this issue fit the bill very nicely in terms of their high quality and industry relevance.
We are still in the planning stages for the last issue of 2007, and are also determining what themes are important to highlight in 2008. I certainly have my own biases, but rather than outline those here, I would like to hear opinions from you, the Micro readership. What topics are most important for us to cover in future issues? What emphasis should we place on near-term or long-term research, or on actual industry and academic designs? Is the technical content and length of our articles appropriate? Any other ideas to share? I always welcome your feedback at firstname.lastname@example.org.
David H. Albonesi
Editor in Chief