Position Statement for DATC Chair-Elect

The hardware and semiconductor industry will be facing considerable headwind in the coming months, and as a result the design automation industry will encounter great uncertainties. However, I would submit that these challenging times are also the most interesting ones and present a golden opportunity for design automation to establish a central and indispensable position in the design and technology development process. Design Automation is largely about productivity and enablement; how to do more with less is precisely the right medicine to encourage the hardware industry to take calculated risk to invest and lift the industry out of the current recession. My agenda for the DATC will be to provide a clear design automation roadmap for the near and long term with a keen eye on expansion into new areas, focus on a holistic view of design tools and the design process for maximal productivity gain, and drive a collaborative model to facilitate research in an adversarial economic milieu.

The high level themes of my agenda are integration and collaboration. To understand what to integrate and what to collaborate on, we need a clear vision of the future. In a time when the DA research budget across industry and academia is reduced, the DATC should step up to the challenge and provide a long term DA strategy. I support the roadmap effort of the base technology, designs, and systems, but propose to explore emerging and new areas for potential expansion such as design automation for biotech and pharmaceutical applications, Design for Environment tools, DA techniques to facilitate the synergy between chip and system architecture and the software stack. I propose to set up a task force to explore whether and how design automation should expand into these adjacent spaces.

During my tenure as design automation strategist at IBM, the main de-railers to productivity I have witnessed are compartmentalization and the lack of a global view. The value is not in treating each component in the design process as separate entities and optimize each one within its individual silo. There is tremendous gain in developing an efficient design methodology by streamlining, modularization and convergence. Therefore, I support the work of the EDPS sub-committee to raise the awareness of design tools researchers with respect to how their tool fit in the overall methodology. Moreover, with each technology advancement, different components within the design process interact with each other more closely. For example, the performance many-core architecture is highly sensitive to the physical floor-plan, cell layout impacts manufacturability and yield, many-core chips are hard to program, and so on. I propose to have the sub-committees to focus on the interaction and possible integration of components that have tight coupling with each other.

Funding for DA research, in both academia and industry will be increasingly challenged. We must go beyond traditional funding models and seek creative ways to extract value from collaborations. A win-win strategy for academia and industry is to create long-term and extensive internship programs so that students can get support for working on exploratory but industry related work and industry can benefit from harvesting top talents and reduce overhead and stretch investment dollars at the same time. Faculty members can also leverage such programs to form close collaborations with industry. The case for this type of “collaboratories” is even more compelling if government money is available to sweeten the pot (e.g. the Taiwan government currently funds such internship programs). The main hurdles to overcome are IP ownership issues and access of proprietary data. I propose to set up a sub-committee on collaboration which draws on membership from leading academic and industrial organizations to focus on providing recommendation for these issues and set up several of these prototype collaborations by end of 2009.

Conferences and workshops play a vital role in facilitating collaboration among attendees, disseminating research, and spurring innovation. The economic downturn has threatened the well-being of some of the DA conferences. We must continue to work with CEDA to revitalize and influence existing key conferences through sponsorship and oversight, and seed new workshops to explore new frontiers.

I am passionate about the future of Design Automation and I believe that I have an actionable plan to work with CEDA to drive the agenda. Thank you for your consideration.