Yes, great question. Good question, Steve. So yes, we view the IoT, it's the kind of latest marketing buzz in the world of semiconductors today. But we view IoT as really an extension of the things we already do. So there's a need for precise sensing, precise signal conditioning and conversion at ultra-low power levels. And there's -- our customers are asking us for -- to be able to take some of the bits we generate and develop information to interpret the bits and give them information, which we do in areas, for example, like reliable health care monitoring. So I think the way to view it is we've got a platform of technologies that we apply to industrial applications, take industrial automation, for example. So we've got the conversion technology, signal conditioning, amplification. We've got processing both on the DSP side. We've fixed-function processing that we develop. We're quite a larger user of ARM technologies, for example. A lot of our products now use MCUs, very, very low-power MCUs which we tailor for our applications. We've the RF technologies that we can bring to bear for the connectivity. And obviously, power management is a core part of the overall solution as well. So the underlying foundation of silicon technologies will be supplemented over time with more algorithmic technology. And Lyric, for example, that we acquired several years ago, is a key player in enabling us to bring algorithmic technology to bear. We're looking at how we make each -- anything that gets connected needs security. So as connectivity in the IoT sector is synonymous, it needs security. So that's something that we're actually experimenting with. We're looking at possibilities to acquire some technology there as well. So that's -- what I've described to you is we've got a very, very broad base of technologies that are enabling ADI in our traditional markets, these more connected or IoT markets, where sensing, measuring and interpreting is very, very important. And we're looking for tuck-in technologies and organic developments to move ADI further up the stack, for example, into the interpretation using algorithmic technology as well as security and communications. So hopefully that helps you.