The EDA Consortium is a key sponsor of the Design Automation Conference
One of the big new things that the EDA Consortium is doing at DAC is a "Productivity & Innovation Luncheon" that will take place on Tuesday, June 5, at noon.
The EDA Consortium was formed in 1989. Its mission is to promote the health of the EDA industry, and to increase awareness of the crucial role EDA plays in today's global economy. To this end it seeks to identify and address issues that are common to its members and the customer community its members serve. The Consortium represents over 100 companies in the EDA industry. It is governed by a nine member Board of Directors, each elected for a renewable two year term.
The Consortium presently engages in the following major programs:
At this year's DAC, the Consortium will co-sponsor the Executive Briefing held on Sunday, June 3rd. This DAC kickoff event provides a status of the industry as well as a business and technology forecast for the coming months. On Tuesday it will host a Productivity and Innovation Luncheon to examine the complex factors affecting engineering productivity, including such topics as communications between engineers and their managers, and productivity benchmarking. The Hacks and Flacks roundtable, now in its fifth year, provides a forum to journalists, analysts and Public Relations professionals to address current communication issues and methods.
The EDA Consortium is the place to network in the EDA industry and address industry issues and opportunities. For more information, visit www.edac.org.
ACM/SIGDA is an active sponsor of DAC
The Association for Computing Machinery (ACM), founded in 1947, has over 80,000 members worldwide and has been a sponsor of DAC from its inception through ACM's Special Interest Group on Design Automation (SIGDA). During the conference, SIGDA sponsors three events that provide opportunities for students and academics to communicate with EDA professionals. The interactions provide valuable input for both sides of the discussions.Ê They guide students in their choices of research topics and career planning, and they give EDA professional a close look at the most promising research areas pursued by universities worldwide.
The SIGDA/DAC University Booth
The Ph.D. Forum at DAC
Design Automation Summer School at DAC
For more information about these and other SIGDA programs, please check the SIGDA website (http://www.sigda.org) or contact SIGDA Chair,
Advise & Consent: Assembling the Technical Program at DAC
by Peggy Aycinena
In a world of constant change, some things stand the test of time. Hence, it's February and once again the Technical Program Committee for DAC is hard at work sorting through hundreds of paper submissions, intent on assembling a robust program that covers the waterfront of topics and technologies related to design automation. The technical sessions sit at the heart of DAC, so this work is crucial to the long-term success and viability of the conference.
As always, assembling the program for DAC is a group effort from start to finish. This year's team of 80 committee members is headed up by Sachin Sapatnekar, professor of ECE at the University of Minnesota, and Leon Stok, director of electronic design automation in IBM's Systems & Technology Group.
I spoke with Leon and Sachin recently -- Leon in East Fishkill, New York, and Sachin on sabbatical in Barcelona, Spain -- and came away impressed with the dedication and endless energy that this entire process entails.
Leon told me, "There are indeed a few points in time when this is a lot of work. The week of Thanksgiving, in particular, is a long one for the co-chairs. Sachin and I personally sorted through all the submissions -- over 750 -- during that week in order to assign them to the appropriate subcommittees. As our areas of expertise are very complementary, I believe we are confident that we made the right decisions there."
"My area of expertise has evolved over the course of my career. I started with high-level synthesis in my Ph.D. thesis, and then worked on IBM's Booledozer synthesis product. I then worked on physical synthesis, and also worked in routing and DFM-related issues. So, I've gone from high-level design right down to manufacturing over the last 15 years."
"Sachin's expertise is complementary to mine in that he has put quite a large focus on timing analysis and circuit optimization. He comes to this effort with a deep knowledge of power and physical design. He also has the academic perspective, and a view of the emerging topics that include quantum computing, synthetic biology and nanotechnology."
How many of the DAC 2007 paper submissions fall into each of the major topic categories? Sachin and Leon told me there are four main categories and gave me the distribution. Topics related to design tools constituted approximately 30 percent of the 2007 submissions, down from 37 percent for DAC 2006. Design methods rose slightly from 38 percent in 2006, to 39 percent in 2007. Embedded systems submissions stand at approximately 15 percent for 2007, while submissions related to design case studies rose to approximately match that at 15 percent as well.
Leon said, "We've also seen a good increase in system-level design papers, so that pipeline is definitely starting to be fuller. That's good because people have seen the need for a long time to build better system-level tools. It's pleasant to see this uptick because there are some excellent contributions in that group of submissions. Of course, we need to connect system-level design to the implementation reality. The big productivity boost we're all looking for simply will not come without this direct path to implementation -- something that's a lot easier said than done. A lot of focus at DAC 2007 and in the coming years will be on how to build that bridge going into the future."
Sachin said, "In addition, we have a number of submissions on near-term and longer-term emerging technologies, ranging from 3D integrated circuits to FinFETs to nanowires/nanotubes to quantum circuits. Traditional areas of EDA are also strongly represented in the submissions, with new twists that reflect technological challenges. For example, several submitted papers on logic synthesis, physical design, and interconnect analysis tackle the challenges of current and future technologies, including nanoscale effects such as reliability and leakage, issues related to developing scalable methodologies, and algorithms that can tackle increasingly large problem instances. Analog, mixed-signal and RF design, and CAD techniques, as well as test techniques in nanometer technologies have drawn a great deal of excitement."
Leon added, "In other areas, we're also seeing a good influx of DFM papers. There's still a lot of ground being broken there, with some new ideas surfacing from academia and some from industry including quite a bit of startup activity."
"In the area of DFM, the ideas are only going to make a difference if you can affect the shapes being designed, and that's not something that can happen on the side. DFM is not a point-tool friendly space because the interoperability of the platforms for DFM is crucial for its usefulness to be realized. So, you're going to see a couple of good DFM ideas that will be picked up and implemented in the mainstream flows from the big vendors. Of course, not all of the small vendors would agree with me here, but I'm only observing this from the outside. Certainly there are some who would argue that if you can produce enough improvement with a DFM point tool solution in the flow, people will pay attention. Those ideas are represented in the paper submissions under consideration, so at DAC in San Diego, we'll continue to hear from both sides of this argument."
Both Leon and Sachin are extremely enthused about a new topic category being introduced in San Diego at DAC 2007. That's the Wild & Crazy Idea (WACI) category. Leon said, "People were invited to submit early ideas, ideas that have not necessarily been tested. They were told that such ideas should be promising and should benefit from the exposure and feedback they might get from being showcased at DAC."
He added, "We received 54 papers in this category and have found ourselves pleasantly surprised that people reacted so adventurously. People clearly were delighted to step up and offer new, untested ideas in hopes of having a hearing in San Diego. We're not sure at this point how many of the 54 papers will be in the final program, but we hope to assemble a session or two consisting of a series of short papers."
Are there risks of potentially patentable ideas being "leaked" too early into the community through this process? Leon said no: "First, just like for regular papers people may already have a patent request in place even before the idea is being showcased in the WACI session. In addition, in the USA, you have one year from public disclosure to submit a patent application, so people have time to submit even if they present at DAC. Plus, these sessions at DAC will provide a publicly recorded event with the author's name on the idea. The presentations will not preclude the author laying claim to the idea, and may even enhance that claim."
I asked the co-chairs what kind of amusement metric might have been met among the many WACI submissions. Leon said, "Well, I certainly have been amused personally by a number of the papers I looked at!"
Sachin added, "This is the first year for WACI, and it has been a learning process for both the authors and for the committee that selects the papers. The whole idea of the WACI track is to be extraordinarily innovative -- but how wacky can you be, while still being taken seriously? We're trying to find the right balance. The wackiness quotient of our submissions runs along the spectrum from conservative to downright crazy. That's good, because it gives us a nice set of choices for the program."
Although both Leon and Sachin have been hard at work for many months now in their role as co-chairs of the DAC Technical Program Committee, clearly there's been some enjoyment along the way.
Meanwhile, it's important to remember that the DAC program selection process only just begins in the fall with the initial screening work. Throughout the months of December and January, the dozens of subcommittee members then take up the task of reading, evaluating, judging, and scoring the various papers to which they are assigned. It's hard, grueling work that can only be described as a labor of love.
The review work culminates in early February at a mandatory, all-hands-on-deck meeting in Westminster, Colorado, where the final program is assembled. The subcommittee members know in advance that attendance is not optional at this meeting. Leon said, "Short of extenuating personal circumstances, essentially all 80 committee members are there in person."
Sachin said, "An average reviewer reads 25 to 30 papers, and we have 80 program committee members who have given significant chunks of their winter breaks to reading and evaluating submissions, all on a purely volunteer basis. We are grateful to them for being generous with their time, and to their families who "release" them for this effort. Nor does the effort stop there. In many cases, our committee members travel long distances to make it to Colorado. For instance, one-fourth of the committee is from Asia or Europe this year. These long-distance travelers may have to deal with uncertainties, such as a recent last-minute disruption in air travel. Nevertheless, people find a way to work around these issues to be present at the meeting, because it's important enough to them."
Leon added, "Overall, final paper selection is a methodical three-step process. The co-chairs do the initial sort, the lion's share of the work is performed by the subcommittee in December and January, and the culminating meeting in Colorado is then an intense 2-day effort that gives us all a chance for lots of good, face-to-face discussion. The subcommittees meet individually at separate tables throughout the course of the first morning, and then come back together mid-day for further discussions. The second day is the culmination and final selection. The conversations in Colorado are substantive and very interesting. Sachin and I have a singular role at this meeting, and it's not one of micromanaging. It is, instead, one of advising and reminding the many people in attendance how they are to proceed in a judging process that leads to a correct and final selection of content for DAC."
"We feel there are two criteria that need to be met. First, of course, the absolute quality of the paper must be taken into account. The second criterion is that we should provide reasonable and appropriate coverage for all of the major topics within the various areas within design automation. We don't want, for example, that special sessions should just address packaging, or just address system-level design, and so forth. We need the DAC program to include a breadth of topics that adequately reflects the reality within the industry."
And what happens if someone brings a private agenda to their committee work in Colorado, or anywhere else along the review process? Leon said, "We feel that the people involved are all technically trustworthy and committed to judging each submission on its technical merit. But we go farther than that to guarantee impartiality."
Leon said, "All committee members are prohibited from seeing or judging papers from their own organizations, whether from industry or academia. Nor are they allowed to discuss those papers in Colorado. In fact, they are asked to step outside when those papers, or their own papers, are being discussed. Additionally, each paper is assigned a 'shepherd' who will watch over that paper as it makes its way through the evaluation process. The shepherd can never be the author of the paper, or someone who is from the organization submitting that paper. We are very insistent on all of this and feel it is an effective mechanism for weeding out conflict of interest."
"An important part of the task that Sachin and I performed in our initial sort was to be sure to get the right paper into the right subcommittee. This is crucial to getting each paper off to a good start in its evaluation trajectory. But, you do have to remember that there are not thousands of experts in each of these niche technologies. In fact, there are only a few hundred, and in some cases there maybe less than 10 people with sufficient expertise to evaluate a certain paper. In that case, we need to monitor the process closely to be sure that a level of impartiality is maintained throughout the selection process."
Sachin added, "We've built a number of checks and balances into the system. The number of papers that a committee member can submit to the regular program is restricted. In many cases, these papers may go to an entirely different subcommittee, and the member may not know where the paper is. Such a policy helps avoid even the appearance of impropriety.
Leon and Sachin reminded me that the DAC paper selection process has been refined over time. Above and beyond the 80 committee members, there is also a legion of external reviewers who are asked to review the submissions as well. These people do not actually sit on the committees, but are experts nonetheless in a particular area and provide valuable insight into the applicability and value of each submission. The external experts help committee members prune and focus the content and help to shape the program into its final form.
In addition to the external reviews, there is also a clarification process that has been developed over the years by the Technical Program Committee. Leon said, "Reviewers can post one clarification question to any particular author, and that person has one week to submit a response. If the reviewer doesn't somehow get 'it' -- the point of a paper -- the author thus has an opportunity to submit a one-page clarification. If the paper should ultimately be accepted in to the final program, that clarification will be worked into the final draft of the paper that actually gets printed in the proceedings."
Is all of this a sufficiently airtight process as to guarantee that great papers never get overlooked? Leon said, "Hopefully yes. Luckily, however, there are other conferences in this space. DAC is, of course, the major showcase for ideas in design automation, so we want to continue to do as good a job as possible in selecting our content."
He added, "It also helps that a lot of the process has become fairly well automated over the years. MP Associates (Editor's note: MP Associates is DAC's Management Company) has procedures in place for collecting the evaluation data and tracking the various scores. That adds to the efficiency of the final decision-making process."
Sachin concluded, "With 750 submissions and 80 committee members, the DAC paper selection process is not a trivial task, but the Technical Program Committee, with the help of MP Associates, works heroically to make things run smoothly. In the final analysis, we are only as good as the efforts of our volunteers, and we have been exceptionally fortunate to work with a dedicated and generous set of folks, who uphold a tradition that is now in its 44th year."
Peggy Aycinena is Editor of EDA Confidential and a Contributing Editor to EDA Weekly.
DAC is Seeking Submissions for the Following:
Women have made important contributions and strides in the EDA industry for over 20 years. In an effort to recognize those who have dedicated time towards these achievements, the DAC Executive Committee presents an annual award to honor an individual who has made significant contributions in helping women advance in the field of EDA technology.
Nominees should have had a leadership role in one or more of the following:
Deadline: Monday, March 5, 2007.