 I'm a transistor guy, a device design that's electrons and holes and it's very mysterious to a lot of people. Over my career, we've gone from, say, 64K to two billions of RAM and two and a half micron technology, but now we're talking five nanometers. Every generation, the chip is half the size, or you've got twice as much stuff on the same size chip, more typically, and it goes faster and it burns less power, and it's the whole thing about Moore's law is it looks like one curve, but it's a whole series of completely new and different innovations at each time. There's got to be a paradigm shift. We need a new device architecture in order to get to the next step. Imagine what the chip can do if it has four times as much power in it. Let's say compared to a 10 nanometer chip, something that's in your phone right now. So you get to that point, what do you do next? And our conclusion was looking at all the numbers and all the simulation and projecting. We need to do something radically different, and that's where a five nanometer comes from. This was a big barrier for designers. EUV, extreme UV, ultraviolet, it has the capability of directly imaging really tiny images. With a nanosheet, you're back to a place where the designer can now use the tools that they have in their toolbox in the past, and these are all really important innovations. We're learning more about how to utilize exactly the same transistor technology to produce a better computing option. So IBM has an absolute need to keep at the leading edge of silicon technology to enable these artificial intelligence and augmented reality kind of applications. So you're not going to get four times as much stuff. You're probably going to get 10. Our interest is to enable this technology to the world. That attracts the partners who are more directed towards ultimately manufacturing this stuff. They come here because they see the value in IBM research. So this concept is on its way out to the rest of the world, and of course we're already working on what's next.