Chasing the wave on EDA research

Many voices raise recently that EDA (or VLSI CAD) is becoming a fading research area, all easy tasks have been solved and hard bones are still left untouched (EDA researchers are smart, they don't touch things like P=?NP). Things are getting worse when the future of the entire semiconductor industry are doomed by the uncertainties of the technology scaling. To keep straight publication however things are hopeless for EDA researchers, we've got to have something new (solving old problems with new ways or "creating" new problems.... I know it sounds pathetic...).

Exotic topics like bio and mobile start to show up in the conventional EDA conference like DAC and ICCAD. Yep, people like to chase the wave of new and hot stuffs, partly because that's where the big money coming from. Although being told by my Ph.D. advisor -- "don't chase the wave, stick with what you're good at", I'm a big fan of wave chaser, actually more like surfing on the front-wave of the new concepts. That keeps me awake and less bored.

This time we took one of the hottest technology -- Cloud Computing -- in EDA. We solved an old problem in logic synthesis -- Boolean matching -- using SaaS 2.0. We do have proposed a couple of interesting engineering ideas in this work, like Bloom filter and cache optimization in the key-value-based database.

For further reading, please refer to the following paper:

Chun Zhang, Yu Hu, Lei He, Lingli Wang and Jiarong Tong, Engineering a Scalable Boolean Matching Based on EDA SaaS 2.0, ICCAD 2010.