The following is a special contribution to this blog from Patrick H. Madden, Associate Professor of Computer Science at Binghamton University. In this blog entry Patrick highlights the CCC-sponsored visioning activity on Extreme Scale Design Automation; the first of a series of three workshops on the topic will be held next week.
The CCC-sponsored Workshop for Extreme Scale Design Automation (ESDA) convenes Thursday and Friday of next week (March 7-8) at the University of Pittsburgh. A group of about 35 academic and industry participants, and many observers from the National Science Foundation (NSF) and elsewhere, will discuss a set of challenges facing the design automation research community.
Over the past decade, the difficulty of designing electronic circuits has continued to grow; the problem has never been easy, but recent fundamental shifts have raised the stakes. First, physical limits seem to have been reached for serial performance; new designs must harness parallel implementations, introducing a host of new timing and synchronization constraints. Lack of gains in serial performance also mean that the software tools and algorithms used must fundamentally change. Further, the abstractions used to simplify the modeling of electronic circuit elements have broken down, and no longer capture manufacturing constraints adequately.
At a time when the problems have become much harder, a shortage of research funding has made academic work in the area more difficult. The problems are too complex for small groups to work on, and limited funds mean that large groups are not practical. Further, the problems faced in designing the next generation of circuitry are so complex that they can drive off all but the most dedicated students – and the best and brightest students available frequently opt for greener pastures. These changes have a long term impact on the electronics industry – we face a shortage of talent for the next generation of electronic design automation professionals.
The workshop series is the culmination of a number of discussions held by the ACM Special Interest Group on Design Automation board; it was clear that a “status quo” approach would lead to a bleak future. The goal of the workshop series is to bring together academics, industry researchers, and representatives from government agencies, to chart the best way to work together and address the challenges.
Each workshop features keynote speakers who will frame the key challenges, followed by breakout group and open discussion. A final report will summarize the findings of the workshop series.
