Frontier Journal       

Exclusive Frontier Coverage on System Design              Vol. 3 No. 6 June 2006

            Kluwer Academic Publisher Editorial - Why the System Design Community Needs the Design Automation Conference and Why DAC Needs the System Design Community Grant Martin

ARM Executive Column - Reducing Costs While Increasing

System Functionality ~ Can Java Offer the Right Blend? Lance Howarth

Acer Executive Column - Acer's Three Major Strategies Stan Shih

Orit Hazzan＊s Column 每 How Does Software Intangibility Influence Software Development Processes?

Analog Devices Column 每 Wireless Short-Range Devices: Designing a Global License-Free System for Frequencies Less Than 1 GHz Austin Harney et al.

TransEDA Column 每 Verification Methodology Manual (Chapter 12) David Dempsted at al.

Morgan Kaufmann＊s Featured Book Chapter Column 每 VLSI Test Principle and Architecture Yinhua Min et al.

Why the System Design Community Needs the Design Automation Conference and Why DAC Needs the System Design Community

Grant Martin , Chief Scientist, Tensilica, Inc.

Many years ago, during the height of the internet boom and the late 1990＊s technology bubble, I began to wonder about the future role of technical conferences and trade shows. In particular, at that time, I participated in some of the ※on-line§ conferences and webcasts that were beginning to emerge 每 both as a speaker and as an ※attendee§. These were interesting times and the experiments with on-line forums for spreading technical information and for discussing products were very worthwhile 每 both to illustrate what was possible with the technology, and to illustrate the limits of the experience.

Despite some reasonable success that I observed with some of these online conferences, they all seemed to have disappeared. We still have product webcasts 每 essentially marketing presentations, often with a chance for watchers to pose questions and for the speaker to gauge some level of feedback, but the main industry and academic technical conferences and trade shows have continued, and the ※online conferences§ have vanished.

Although some of the conferences, workshops and symposia have disappeared over the last few years, for every one that has disappeared, it seems at times that two or more have sprung up to replace it. Often the disappearance has more to do with the dwindling interest in a particular technical area that seems to have become mostly ※solved§, than to do with the desire of attendees to learn, and of speakers to teach. This is in stark contrast to the electronics trade press, for example, which has lost a number of publications in recent years, or has seen those surviving shrink in size, to be replaced in large part with information and misinformation posted on the web.

Why has the web seemed to be a partial substitute for industry information in print, and yet failed to supercede the technical conference and trade show? I suggest that the primary reason is ※people§.

Although information can be obtained in multiple forms 每 reading articles, watching webcasts, browsing the web, reading books and magazines 每 one of the most important ways of receiving and giving real information is simply to meet people in person and talk to them, listen to them, participate in a group listening to a talk, asking questions, and meeting total strangers in the hallways, over coffee, wandering an exhibit floor, and at parties and special events.

How is this relevant to the Design Automation Conference (DAC)? Quite simple 每 as the leading worldwide technical conference and trade show in the area of electronic design automation (EDA), DAC plays a leading role in building the community of design and EDA technology professionals 每 both in industry and the academic domain 每 that now spans the world. In addition, by being a forum for real teaching and real learning 每 in the technical sessions, from panels, special sessions, industry events, and on the trade show floor 每 DAC grows the design and EDA community by giving students a chance to both learn and demonstrate their own knowledge and accomplishments, and by giving newly hired members of the community a chance to learn about the technology and industry far beyond the scope of their own job.

As technical co-chair (methods) for DAC 2006, which will be held in San Francisco from July 23 to 28, I have worked with my co-chair, the entire technical programme committee, others involved in DAC planning and those responsible for pulling the entire conference together, to make sure DAC 2006 has a very strong technical programme. This year, we feel that it may be one of our strongest programmes yet.

But having a strong programme means very little if people don＊t come. Our hope is for increased attendance and a wider participation from the design community 每 including elements of the community that may not have come to DAC before, who find some exciting new content in the programme (please check www.dac.com for details). Again, the ※human factor§ is the only real factor that makes a conference successful.

How is this then relevant to systems design, and the systems design community that this journal is written for? I would argue that in particular, systems design is not yet a ※community§, in the sense that, for example, the ASIC design community is a relatively coherent body despite being spread throughout the world. The system design world is more of a ※community of communities§ 每 with many disparate and heterogeneous niche groups who perhaps have more in common within the group than between them. This is of course a byproduct of the heterogeneous nature of systems design and the systems products themselves. DAC 2006 reflects this heterogeneity in its system and embedded oriented sessions 每 including those concentrating on embedded software, secure systems, the design of multimedia, entertainment and games, system diagnosis and debugging, multiprocessor system-on-chip, networks on chip, links between system level and verification, and transaction-level modeling. There is a workshop on UML and SoC, and tutorials on hybrid modeling (applied to domains such as automotive and avionics, for example) and SystemC. There are also sessions dealing with areas of strong interest to some of the system design community 每 such as ※Beyond the Die§, - chip/package codesign, higher level interconnect and analysis 每 as well as introductions to new and emerging technologies such as nanotechnologies and lab-on-chip.

On the exhibit floor, as well as a large number of EDA companies, there are companies involved in intellectual property based design and other areas of system design.

With such a heterogeneous array of offerings, it is easy to see that system design at DAC 2006 brings together many communities into a common place over a common time. The opportunity for people from these heterogeneous system communities to break down walls and barriers between them, to learn things beyond their own scope, to meet and discuss all manner of technology issues in the hallways and coffee breaks and on the exhibit floor, is an opportunity for DAC to help build a systems design community that begins to be a more integrated whole. Systems design needs to find venues to help it come together, and DAC 2006 provides an excellent place for this to happen.

We can see why the systems design community needs DAC. Why does DAC need the system design community? Simply put, systems design (whether under the name of ※SLD§ or ※ESL§ or another acronym) represents a powerful growth area for design automation. Whether created by commercial EDA vendors or universities, by systems companies or by semiconductor companies, by large firms or startups, more and more designers need the system design tools, flows and methodologies emerging from all these sources. As DAC is the premier worldwide focus for design and EDA, it must grow to encompass all relevant new communities, and serve their needs be growing its technical focus as well. Whatever designers need in the way of design aids, design automation, tools and methods in the electronics and embedded system and semiconductor domains, DAC exists to serve. Growing DAC to be of more service to our system design communities is a next big step.

I hope to see you all at DAC 2006. If you happen to see me, please come up and say hello! It＊s the human touch that builds our community, after all.