IEDM 2009 preview (Dec 2009)
Its been a long time between posts which I blame on the swine flu, but enough of excuses. In less than a week, leading semiconductor technologists from around the world will gather in Baltimore for this years International Electron Devices Meeting. The venue change is not related to the economy and lower expectations for attendance. The traditional conference hotel is being renovated. But like just about everything except unemployment and bankruptcies this year, I expect the turnout will be down from previous years. Thats a shame considering the quality of the papers that IEDM attracts and the hours of work that went into most of the research being presented.
A good plenary session seems to set the tone for me. For my money, this has always been the highlight. It is the session where you hope to listen to a few real visionaries. Too often, we think of someone with vision as simply looking past the next round of quarterly financials. IEDM attendees have come to expect more than that. This year, I think the three plenary talks fit the bill of looking well past next year.
First up is John Chen from Nvidia. As one former colleague often noted, the graphics processor has too long been considered the poor cousin to the microprocessor. The GPU just does not get the respect it deserves in terms of design sophistication, complexity and processing horsepower. Hopefully, attitudes are changing. At least my Apple computers run software that take full advantage of the GPU to speed things up. Perhaps John Chen will help to improve the image of the graphics chip when he kicks off IEDM with GPU Technology Trends and Future Requirements.
Few would dispute the foresight of contemplating our everyday environments filled with electronics printed onto the surfaces that surround us. In that sense, the second plenary talk exceeds the standard for taking a longer term view of our industry and our world. Takao Someya from The University of Tokyo will present Printed Organic Transistors: Toward Ambient Electronics. This talk seems to establish a trend connecting these talks back to my old Chipworks and SI colleagues. Ray Haythornthwaite wrote a landmark report discussing developments and future trends in organic semiconductor technology almost eight years ago. I remember Ray envisioning electronic wallpaper that would envelope you with your favorite screen saver. (Wouldnt that save us all a few renovation dollars?) He also realized the possibility of creating cheap printed electronics where miniaturization was not an issue such as devices like RFID tags that could be printed as part of a shipping label.
The final plenary has a looser connection to former colleagues (although a couple of them have done extensive failure, reliability, and IP work on implantable electronics), but one is hard pressed these days to get far from the debate about health care considering US President Obamas plans for reforming the US insurance system. In that sense the third presentation is well positioned. Given the health care crises faced by the developed world, this talk details some innovations that will at least slow the rate of cost increases and hopefully have the audacity to actually reduce the financial burden of medical care. Perhaps if the costs are low enough, some of the technologies discussed might play a role in the developing world as well.
The IEDM press kit suggests one hot item to watch. There is a novel memory that works by distorting a fin with electrostatic force so that it is either centered between two electrodes or contacting an electrode. Who needs all that electronic and magnetic switching (not to mention phase changes)? Researchers from KAIST University in Korea decided to make a good old fashioned mechanical switch. Their device integrates NEMS switches and CMOS drivers and readout circuits. (If youve heard of MEMS, NEMS is just the updated form now that we have passed the micro generation and are now into the nano.) Of course, this technology is for demonstration purposes only, but its always good to see what is coming out of research institutions. Sometimes, there is a tendency to discount technology that is not clearly ready for commercialization within months as opposed to years. But my thoughts on the rob your future to make a quick buck mentality needs its own time and place for a full discussion. See Paper #27.2 at 9:30am Wednesday, December 9.
It may be yet another promise that will go unfulfilled (references to political figures living or dead is completely unintentional), but lets think of this post as the first preview or a pre-preview of IEDM with more to come later. My verbosity has overwhelmed even me with this exhausting look at only the plenary and press kit. I hope to cover the breakout sessions in the coming days.
Check back soon.
Tuesday, December 1, 2009