Spent much of the day on the show floor, but started off with an outdoor briefing from Mark Cieri of Enperion, the Lucent spinoff with the tiny, fast-switching dc-dc modules that incorporate the magnetics in the package. We talked about the growing demand for point of load regulation for DDR, something that I hadn't thought about. The dominant technology today is still LDOs, but for new blade designs, once you've maximized efficiency in powering the processors, memory is the next obvious target. Curtis and I did a short video of the interview, and it will probably be up by the time this blog entry is posted.
Speaking of videos, there is a neat vid that we shot at the Intersil booth, where you can watch a self-calibrating point-of-load calibrate itself. Well, you can watch the process in the time domain on a screen. The point of self-compensation is that in manufacturing, every board is just a little different, and one size doesn't fit all. The point of being able to watch it happen on a GUI is that while you're setting up the production line, you can watch verify what's going on inside the package:
I also got a video interview with Carl Blake at the Transphorm booth. The company's not exactly announcing pricing. Carl explains what the process of engaging with Transphorm is. There's also a demo on the video.
I had a briefing with TI that is not on video. Got some advanced information and an explanation of how TI's new driver for GaN FETs works. There are some particular requirements for turning GaN FETs on and off. That's something I'll be writing about in Electronic Design in print and on-line.
After checking out the tradeshow, the news is that there is not some new power paradigm stalking the aisles. On the other hand, this is a very well attended show, at a venue that it is not cheap to send employees to. I'm taking that as a mildly hopeful sign for the economy in general and for power in particular.