 So here at the Computex 2015. So what are you launching here? We are launching two things. We're launching a radio IP which is what you see in this chip that I have in my hand as well as a new subsystem for IoT. So radio IP. So where is this IP? The IP is actually as you can see the chip here. It's actually embedded inside that chip. This is a reference design and the application that we are demonstrating with this design is called Beacons. Sometimes you might hear the term near-ables as well. So those Beacons are based on Bluetooth LE technology and the radio IP does the Bluetooth LE protocol and it's inside that chip. So it's ARM Cordio? It's ARM Cordio. That's the name of the IP. Cordio IP. Is it just Bluetooth 4.2? Is that what it is or it can be different things? It's actually Bluetooth LE. So it's the low-energy version and it's fully compliant with the 4.2 version of that spec. So is this digital analog? What is it? It's both. It's got actually it's got an analog component, which is what the RF is done with. It's got the baseband. The modem bit is done in digital, but there's also software running on it. So the control for that Bluetooth, the link layer control is also inside there. So this is not Cortex-M stuff. It's not Cortex-A stuff. It's something different. Actually the software runs on a Cortex-M0 plus that is embedded inside that IP. All right. So how about the software? What's going on with the software right here? Okay. So the other thing that we launched today is the IoT subsystem, as I was saying, and that IoT subsystem is based on a Cortex-M3 and that's designed to run embedded OS that Chris can talk about. All right. So hello. So who are you? I'm Chris Faulkner. So what were you talking about on stage? Well, we were talking about a lot of things. All the building blocks to build IoT devices. And I guess the bit that my group does is actually the software side of this. So it's Embed OS and Embed Device Server, which allows you to build a solution around these types of devices and then connect it to the web. What's the Embed Device Server? Embed Device Server is the cloud component of our software. It basically allows you to take an Embed OS-based device and then plug it into a cloud and the data flows from the client to the server and you can also manage these devices through it. So when there's 50 billion devices in the world, how many of them can be this kind of solution and embed? Then what do you think? Well, good question. I mean, if there are other 50 billion IoT devices, a lot of them are going to be small sensor-based devices and also battery-powered. What fraction? I don't know. But I would actually estimate the majority over time will end up being these tiny devices. So this is software for tiny devices, right? Like how small are we talking? Cortex-M, right? This is a Cortex-M class. It's a bit like the 80s all over. When 64 kilogram was seemed like a lot, that's the class of devices. And the flash size is a few hundred Ks to maybe a megabyte or two. That's where they start and they go up from there. But it's that level. It's kind of like ARM 1, ARM 2, ARM 3, stuff like that. Yeah, something like that. It takes me back to my childhood. I mean, that's basically what it is. All right, so you announced this a few months ago and what is the status? Are you releasing it very soon? It's coming to a public release. I believe it's in November. The beta will be slightly before then. Yeah, we've had a number of releases to our partners. We've done three alphas. We're just about to do a beta this summer. Yeah, so it's on track. And how soon is this IP available in products? Well, the IP that I'm showing you on that chip is at a prototype level. But we're now finalizing it. We've actually built our final silicon. That silicon that you saw earlier is not the final silicon, but the final silicon will be coming out very soon. And we are in deep discussion with partners for licensing the IP so that they can go into market at the end of this year early next year. And you had the TSMC as a guest here and you're talking about a new 55 nanometer for this? Yeah, the technology is based on what TSMC calls the 55 ULP process, which is ultra low power. And one of the key benefits of that compared to the traditional LP process is the dramatic reduction in power reduction that you can get. So how low power are you talking about? Is low power doing everything else until now or is that the point? It's all relative, of course, yeah. So it is very low power compared to other 55 nanometer processes. And I don't, I can't remember the numbers off the top of my head. But on my slides, I showed some numbers which give you an idea of the amount of power saving that you can get. The implementations that we've done with our Codex M3 subsystem, we were able to get 25% reduction in the leakage, but also a 40%, more than a 40% reduction in dynamic power consumption.