 Hi, I'm Damian Wells from NXP. Welcome to Embedded World. We'd love to show you some of the really cool demos that we've got going here on the booth today. Hey, Kyle. Can you tell us, gentlemen, what we have here on this particular demo area? Sure, absolutely. Welcome. So at NXP, we're showcasing our microcontrollers in this display. To start, we have our 1170, the crossover microcontrollers. This is a voice recognition example where I can say, hey, Oven. And hey, Oven, shut the door, and it will close the Oven door. Hey, Oven, shut the door. All right. So it's always listening. It's for voice applications. Yes. This does some vision where it can recognize gestures, start some audio, or you can close the audio, or you can use the touch, both showing the human machine interface and the vision and the audio detection. What is a chip that can do this? Yeah, that is the IMX RT-1170. It's a dual-core, high-performance crossover MCU from NXP. It has an M7 and an M4. In this case, we're actually using both cores. The M4 is using more for the human machine interface and some of the general activity, isolating all the complex voice and vision on the M7. And you're launching new chips here at the show? Yeah. So NXP is showing a lot of applications and use cases for the MCX. It's our newest microcontroller. It's a dual-core Cortex M33 device. We have a demonstration here, which is showing the human machine interface with our free-to-use GUI Guider software tool to create rich user interfaces to drive interaction with the customer. And so dual-core M33. Yes. It's good performance for this kind of applications. Absolutely. It's up to 150 MHz per core, but it also has a graphics accelerator. It has a DSP and it has a neural network application for machine learning. How do you do the neural networking on a MCU? So what it has is it has that, it's an NXP neural network core called the Neutron. And we use our EIQ software to train and learn the different machine learning algorithms. We download it to that microcontroller, but it offloads all the machine learning to that assist neural network. And it's part of the core? It's part of the chip? It's part of the same chip, along with the two dual-core Cortex M3s and the hardware accelerator for graphics. And what are you showing here in the laptop? So I'm excited to present the Visual Studio code extension. It's called MCU Expresso for VS Code. We've gotten lots of requests from customers to provide another option for development to complement our IAR, microvision and eclipse-based tools. So now customers can use the Visual Studio code environment to get started developing with NXP libraries as well as a Zephyr development. All right. So it's busy at your booth. A lot of people asking about these new chips. A lot of people are excited about both the vision applications on the RT-1170 and also the MCX and the new microcontroller that NXP has provided customers. Cool. Thanks a lot. It was nice meeting you. Hi, let me take you now over to our matter demo that we have running over here as part of our smart home demonstration area. So Michael, could you explain what we have here in terms of our matter demo? Yes. So this is our matter demo. So this is for the smart home and IoT. The idea of this demo is really to demonstrate the interoperability of this new standard. So we are talking here about a new standard, a new common language on the IoT markets and NXP is a great contributor of this new protocol and we are offering a new platform, complete solution, hardware and software so that you can develop matter devices. So our broad portfolio helps us to provide solutions for gateway and matter controller and thread border router. But we are also providing solutions for matter and device over thread and also matter and device over Wi-Fi. And then there is a controller which is just Android. So yeah, on this demo what we are trying to show is we have several controls. So we have this Android matter controller but we have also a number of matter controller which is available on this demo. What is running here? So here is running the matter controller and the thread border router but you can also control some of the end device if I just walk through here and then you can see that we can control this smart rack and fan directly. But also which is a key point here we have also on this demo, third party controller with iPod from Apple and Nest from Google and this controller are also matter compatible. And what you can just say is, hey Google, turn on the light. And then if I say, hey Siri, turn off the light. So as you can see then you can show that whatever the ecosystem as it is matter compatible, you can drive all your device in your smart home. It's very popular to work on a matter right now? Yes, that's really, this is a very beginning of matter but we see a lot of adoption. It becomes very popular, a lot of buzz around that. But that's for real now and it will expand and we for sure see a lot of promise in matter. Yeah, that's a promise of matter. Cool, thanks a lot. Thank you. Alright, let me take you now over to our UWB Ultra Wideband demo and show you some of that cool technology in action. Hey Lois, do you want to explain what's going on here with UWB in the demo? Yeah, of course. So UWB is a radio frequency technology which allows us to range very, very accurately. We can range 10, plus minus 10 centimeters and at an angle resolution of plus minus 3 degrees. That's also what we're showing here. I have a smart phone from Samsung, just an enormous smart phone with an app for auto wideband and an auto wideband chip in there. And I have another auto wideband device based on the SL-160 from NXP in there. And it's on top of the bezel, it's two antennas and if we step back I can move the smart phone around. And on top there you can see the angle change in relation to where I'm holding the phone in front of the laptop. So the distance changes, if I move back the distance changes and if I move further in again it's again close distance. And what we're showing here with this demo is seamless logical access. So I'll just turn it on and imagine you're working on your laptop and you just move away and you forgot to lock it. And because I am connected via my smartphone with auto wideband to my laptop, now if I go back in to my laptop and want to continue my work it's unlocked again. So this is showing the seamless logical access with auto wideband. What happens if you forget your smartphone? Because if you forget your smartphone and you move away and if you move behind the white line then it's still unlocked. But because what we're showing with this demo, the SL160, try to mention, is that we can do auto wideband radar and distance information at the same time. So now I step back from my laptop and as soon as the laptop does not detect my presence anymore with auto wideband it will lock it. Because there's no one there. It's locked for security reasons. And now if I go back it does detect my presence again but it doesn't of course know that I am the person that is allowed to use this laptop. So I need to put in a pin again and yeah this is just the amazing... As long as somebody is near the laptop it's unlocked. You have to prove yourself that the person is allowed to use the laptop. So it senses people? With auto wideband radar we can sense presence, yes. Presence, yes. And all these support act? Yeah, exactly. We have the try mention SL150 with the module partners Amotech, Murata, SGI, Sunway. And we also have using the SL150 development kits which can all do auto wideband distance measurements. And we also have the SL160 which is a more radar optimized chip. Alright, cool. Thanks a lot. No problem. Have a nice day. Enjoy the show. Cool. So we're going to go now and take a look at our high voltage battery management system demo here with Andreas who will explain how that is and how it's working. Hi, it's my pleasure to explain your high voltage BMS reference design, the cloud connected BMS reference design. So this is what we call the high voltage 400 volt reference design with the main board with the microcontroller and the intelligence, let's say, the software stacks. This on that side is the cell controller, cell monitoring. So each of these small little guys here control and monitor 14 battery cells. And then on that side we have the high power rail. There's a shunt and then there's high current measurement. For redundancy reasons, for safety, functional safety reasons, it's a dual of them. So all of that is we provide to our customers for prototype building together with software stacks and functional safety analysis. And then this demo is actually more, is one part of it. The other part is that we partner with the company Electra Vehicles in. It's actually a demo of two companies. And I have here my friend Fabrizio, the CEO of Electra Vehicles in. And he is doing the cloud and will explain to you the cloud business. Exactly, yes. What we have done in partnership with NXP, we have demonstrated the ability to connect to the cloud and leverage some, not only the edge computing power of NXP chipset, in particular that's 32K3 chipset here, but we also able to connect to the cloud and leverage an advanced digital twin of the batteries. So what is a digital twin? It's basically a copy, an exact copy of the battery, but in the cloud. And what we do at Electra, we basically create a digital twin with the physical base, so actual data coming out from the battery itself and machine learning. So we leverage a lot the latest techniques of machine learning AI in order to have a very accurate digital twin of your batteries. And what we do is basically when there is an error that starts to give some problems to the state of charge or state of health on board, we basically ping in the cloud and say is there a new update, is there a firmware update, in order to download it and reduce the error, make this error disappear. The opportunity is that we are going to have with electric vehicles very accurate state of charge, so we know how many miles, how many kilometers we have left in our vehicles, but also to have a very good state of health, so we know if the battery is healthy or not. And finally, if we can predict potential failures up to three months in advance, if our batteries will go well or not. So we are showcasing here the partnership with NXP and Electra for a very advanced, very novel BMS solution. Because the batteries can be very advanced, they keep coming new batteries all the time and it's very important to manage how fast you charge, how fast you use it on the road and it's a complicated system, right? Absolutely, yeah, batteries are alive systems are very complicated and what we are seeing is that there are new batteries coming out, you know, NMC811 or the novel... What is that one? Which one? What is this you just said? So yeah, there are different chemistries, you know, the popular one by Tesla is NCA, there are many from the Chinese market LFP and so on, but there are new trends like NMC811, they are very advanced chemistries as well as lithium metal, solid state, potential in the future we will see lithium air and it's very important to do a digital twin of the battery in order to have the ability to know the battery itself and be able to deploy advanced algorithm to make any chemistry working well. It's like simulation, emulation, something? Yes, exactly, you basically create a digital twin, so a copy, a digital copy of your battery but in order to do that you need to do a lot of tests, a lot of machine learning but once you have that model you can do a lot of things and you can leverage the network effect. The battery is the most expensive part in an electric vehicle so you want to maintain that over a long lifetime and of course you also want to get the maximum distance out of your vehicle so that's why you need to have very precise algorithms and the digital twin helps you to get these really precise. One thing I want to see, but I don't know if the car makers want to do it is battery swap and that they should be a standard and everybody should swap the batteries but that will mean there will be a million different batteries out there and each of them different health, different range, different... but that's a different question of it. It's a different question, there's companies out there that have tried that in the past there's also companies out there that have tried that now but it is like you say, I mean, vehicle manufacturers have own ideas about their batteries so they don't want one unique battery they want different performance levels for their cars. What do you think? It's a differentiation point. Yeah, I agree with Andreas that we're going to see where our batteries swap in the past we're going to continue to see more and more it's definitely, we need more normalization we need more rules around the batteries but for now it's a very early in the industry and we're going to see a lot of differentiation in the adversities but we might get there, it just takes 20, 30, 40 years of time or so. I don't want to wait so long. I know. I want to wait two or three months. Oh, I see, you want to swap the batteries real quick. Two or three years, okay. I see, okay. I don't know, I think the European car makers need to move fast and do something, I don't know. I don't know what's happening exactly. Sure, but it's a very complex matter and that's why there are companies like NXP, Electra, many other players in the industry try to push electrification forward make sure the battery is safe and they give you the miles and the mileage that everybody wants the batteries swap potential will come, it will take some time not two, three months. Okay, cool. Thank you. Thanks a lot. Thank you very much. Thank you. Cool. Let's get another one of our demos that we have here on the booth which is our concept e-scooter. Nice. So here we have basically a representation of an e-scooter obviously made out of Plexi and we've actually positioned within it many of the different building blocks that NXP could provide to someone who was wanting to design an e-scooter for themselves. This is just a whole bunch of NXP chips that goes in a scooter. Exactly. It's actually combined together with a story that is actually running on a screen that's just off to our right-hand side here and this actually talks about the e-scooter itself in terms of future mobility. We can drive into more technical details around the display connectivity controller, the onboard charger, traction inverter, the battery management system and indeed we can even drill down to the individual parts that NXP provide that actually could be used within each of these applications. So very visual, tends to draw lots of people in who are potentially looking at designing products within that area. Nice. I see a chip down here for the tire pressure, the radar in the back, onboard charger, just a whole bunch. Cool. All right. You have your BMS here for the battery management system, electronics for a handlebar control. We have a display up here which is being driven by a display controller that sits in the front of the bike and of course forward-facing radar as well. Is there a trend of trying to put more and more stuff and less fewer chips somehow that can do more? For sure there is. Obviously everyone is trying to reduce down essentially the amount of electronics that you would put into something like this but they all have pretty dedicated functions as well. So, you know, the BMS is very much its own separate entity and you've already heard about that a little bit on the high voltage one for the cars over there. Can I take you also over to the partner wall that we have here on the booth? Cool. All right. Because NXP has a lot of partners in the embedded world. NXP has somewhere in the region of about 250 to 300 partners in total. And they might all exhibit here at the show. Yes, many, many of them are. We have, actually, within the show we have over 100 demonstrations that are not on our booth they are elsewhere. We have a sample here on this wall of 24 of our demos. This particular side of the wall we have a brand new device called the Ida MX-93 that launches next month and this entire side of the wall is actually all of our embedded board suppliers who make these little modules and they are basically, all of them are based on the Ida MX-93 ready for the part to be launched. So, here we see advanced tech doing client tech. So, even though it's a new chip they've been working with you for a while to make it ready for the market. We engage them in the Alpha program so very early on when we were also engaging with our very big OEMs and by engaging them at that point they're actually able to develop a module in time for when the product reaches full market launch. And they, your partners covered all requirements of the market what people want, like all the different segments, the demand is covered, there's a solution for everything and they're more and more coming. So, what's the performance of the Ida MX-93? That's a high performance core, right? It is a high performance core, yes. There's on board an ML engine as well to obviously help with offloading. It says one or two Cortex A55 plus the M33. Yep, for real time. And then actually on... When we look at this side of the wall this is another selection of our partners. The four that are here are actually all HMI, so this is more graphics and the ability to develop a GUI, you know. And then further along the wall we have some of our Wi-Fi partners and indeed some of our automotive partners as well. So when there's a new Ida MX coming it's a big deal for the industry? It is very much a big deal. We don't just have Ida MX on this side of the wall we also have some of our microcontrollers some of our network processes as well and indeed some of our automotive products. So this is one of our automotive products, the green box. And indeed the new part, another new part that was launched at the show was the S32G3 and this particular partner has already got a board out based upon the S32G3. And the S32G3 is computer vision stuff? It is for the automotive domain. Here's a big SOM. Yep, that's an Ida MX 8M plus as a SOM module. So it's been a busy, busy show? It's been a very busy show, particularly the first two days. Today obviously it's a little bit quieter there's a few more students here which is great to see some young talent coming through learning about the technology that we have on offer. Yeah, so you also talk here about automotive? Maybe you can introduce... Certainly, let me introduce Brian Carlson my colleague here who knows all about our automotive products and what we have. Hi, so Ida and I just walked up and saw you here, great to see you again. Have you seen anything yet? Yeah, we did the booth tour. So you've seen the automotive booth? Yeah, let's go over there. Automotive table. Oh, definitely. You saw the automotive partner? Let's try. Let's go over there because there's a lot of excited things going on in automotive definitely, right? We know about electrification, connected vehicles, autonomous vehicles and everything's moving towards software-defined vehicles. So that's what we're really focused on here. So let me just give you the rundown. So we'll start here. Electrification, of course, is a big trend. OEMs are spending $20-30 billion a piece to create electrification platforms. And so we're right in the mix with battery management, onboard charging. This is for the actual drive of the EV. So we introduced this product initially in Munich back in electronic in November. This is the actual chip right here. That's the S32K39. The nice thing about that is it's an ASL-D rated chip that includes a lockstep pair of Cortex-M7s, two additional M7s, but it has a lot of integration. It has sigma deltas, it has really high resolution, peak resolution of pulse width modulation to control gate drivers. But what's really good about it with the has two dedicated motor control co-processors. This allows us to control very high-speed inverters, which we take the 800-volt, typically 400-800-volt battery voltage. We have gate drivers that we introduced also recently called the GD3162s. And we're able to switch those to be able to do a whole control loop up to 200 kHz for two motors. So this chip, why it's so impressive is being able to drive two electric motors at the same time totally independently. And then we can extend that with the S32Z and E, which have one gigahertz Cortex-R52 cores, which we don't show the part here, just visually we actually won best at show at this show in a better world back in June of last year with the S32E. And it has eight, as I mentioned, R52 cores running at gigahertz. It's kind of the brain of the EV. This is the brain of the electric motor, and it's just optimized for efficiency. We can support SIC, SICK, or Future Gallium Nitride. So it's about high performance electric motor control, the ability to scale from all the way up to four motors for a four-wheel-drive EV. And we have the ability to show visualization of what's going on in the chip. So this is a really exciting device for us. This has been probably one of our most attended demos here. What we announced this week, though, is the production launch of the S32G3. Now, this is a really exciting processor. It actually has 30 processors inside, 21 ARM cores, and it allows us to combine M cores with A53 cores, lock-step support for these, which allows us to do ASLD safety applications. So the two key applications we see for this are the vehicle computer in the new software-defined vehicles. This is actually being used in production vehicles today. We just opened it up so anyone through distribution has access to this technology now through distributors. This is the RDB, or Reference Design Board, Schematics, Layout, Bomb. Everything's included and provided on nxp.com. You can get these boards. And the Gold Box, which is the enclosed version of this, is really popular. So this board actually has 18 CANFD interfaces, 12 Ethernet interfaces up to 2.5 gigabits. It has two M.2 modules here. So we support Halo 8, which is like 27 tops of ML. We can support a terabyte SSD or a Wi-Fi 6. So these are really popular. There's a PCI Express slot here. So it's an extensible SD card. This is a very, very popular product. And you see a lot of Gold Boxes on the floor at partners. They leveraged this for a lot of their demonstrations. And Tier 1's used it in cars to actually drive and do testing of their software with this platform. So this is a production. The big thing here that we're showcasing is the integration of Kubernetes. We have a product called Goviap. It's available on nxp.com. It allows you to put this whole stack with dual Linux running. It has Zen, a hypervisor, and support for Kubernetes K3S. And so what we're really showing here is multiple workloads. Containers running on this board. This is Grafana. We also have Prometheus for event monitoring. But all of that is running on this board. And you can log in through a web browser and actually monitor all these tasks and manage containers in an automotive, software-defined vehicle environment with this product. And this is just the tip. There's a lot of exciting applications that will be coming to vehicles. There's a lot of isolation that allows people to put new applications on it very easily. So software-defined vehicles are a really key turning point in automotive to progress into more smart phones and data centers and be able to deploy workloads very easily onto the job. So it works. The whole idea of software-defined vehicles in the future, you buy an EV, it just keeps getting better. That's the point I always say, right? The worst your car is ever going to be is when you buy it. And it actually gets better over time. So we do this in all these areas. The motor control. We talk about a software-defined motor. We're showing over here the intelligent battery management. We'll see that in a second. Where we're continually updating the machine learning models and improving the rate, getting very, very accurate range and also extending the life of battery 12% to 15%. So just through software and algorithms and machine learning in the cloud, digital twins in the cloud, we're able to combine vehicle data in the cloud and to deploy. And there's the whole development side of this STV. And we're supporting virtualization of the processors in the cloud. So you can actually develop your software in the cloud, do all the testing integration, and then deploy it back to the vehicle through over-the-air updates. And even the Gold VIP has the ability to support over-the-air updates of virtual machines, the complete real-time system, or the complete image. You have a lot of flexibility and really advanced over-the-air updates with intense security. This comes with an integrated, basically integrated secure element. It allows you to do all the public key infrastructure. So it has leading-edge security, leveraging the technology from NXP because we're in secure, you know, bank cards, passports, and leveraging enterprise networking capabilities. It really brings NXP's expertise all into one chip. So this is being deployed by many OEMs. The initial vehicles actually came out in China, Liado. They're already in production of vehicles, and there's a lot of global OEMs that'll be putting vehicles out through this year and next year. So it's the start of the SDV. There's a whole bunch of things going on around it, like zonal architecture and how you address each of the zones. And getting rid of a bunch of boxes and integrating those is virtually easy to use in a box. So we're doing that too. That's what we showed last year with the S32Z and S32U, which are more zonal architecture chips. When you define on software, it's great, but you also sell hardware. So do you want the cars to be able to swap out the board when there's a new chip in the future? Well, that's the key thing about this. I'm glad you asked about that, because we talked about SDV, we talked about the board of reference, the other key message we're talking about here, is scalability, because this part extends our S32G family. We have the S32G, too, which has been in production since May of 2021. What this does, those are four devices. Now we've taken that family forward with the G3. It doesn't supersede it, it actually augments and extends that family. So now we have four new devices. So what we're seeing is we actually have the low-end device of the G2, which is a microcontroller only, with three lockstep M7 cores, and it goes all the way up to this that has the full complement of eight A53 cores and eight, which are four lockstep pairs of M7 cores. So what happens is we have customers that start at the low-end, even the microcontroller version, and they're able to, over the life of the vehicle, to swap out or to increase performance. So if they're doing a certain capability, they can make that smarter, more intelligent over time. So year to year, they may decide they keep the same architecture, they keep the same box. This is pin-for-pin compatible. So all those G2 devices, G3 devices, you can start to add those up over time and within a fleet, you may have low-end and high-end. You could have different versions of the same chip, foot print compatible, or you could extend it into the future to add more capability. Maybe three, four years in the future, I drive in somewhere, somebody swaps all my boards, and then I have the next version of the hardware, but the software continues. Exactly, and that's where Software Define Vehicle allows you to do that because today you may have to replace 16 boxes in the car, which is never going to happen, where you could put one chip or one module and replace that out and double the performance like that. So that's the big advantage, and that's what's new. It's the modularity because a lot of these, like this capability would be part of a module inside of a central computer with infotainment and IVI, or IVI and ADAS in like one central box, and this is the main vehicle computer that works with that, and it's modular typically, so you could upgrade those over time. When people want to optimize the range and the efficiency of the motors, and the regeneration and everything, and you're involved in helping them make that better all the time? Exactly, we have, and we work with a lot of partners, one of our examples. Yeah, we interviewed them just before. So you've already done that one. That's a great example of how you can leverage the vehicle data and the game changer here is the edge processing. To be able to do this real-time, continually, every time the car is running, they'll monitor everything that's going on in the car. That data is really important. We've partnered with cloud partners like AWS, for example, so we support AWS IoT fleet-wise, so we can actually also trigger on certain events. So if there's certain events going on in the car, like a heart-breaking event or some kind of failure, we can actually capture that intelligently at the edge, bring that to the cloud and they can process it across a fleet. So that's where all this data is converged at the cloud. They are able to process it, understand how to improve it, and then they update the models. Just like the lecturer probably told you, they update the models on the vehicle and now it gets better over time. Whether it's the efficiency of the motor, the drive, or the efficiency of the battery, or the charging algorithm may be improved, especially you want to improve the algorithm, so you're not stressing the battery during charging, so that can be improved. There's a lot of opportunity and electrification to make these things longer, better, and last longer. In EVs, there's pretty much like supercomputers. That's what it is. This is a supercomputer. Big performance, and there's also a whole bunch of other processors that are wrong, right? Little processors. You target different clock. We optimize the solution like this is M7 based. This is Cortex A53. Those are Cortex R cores in electrification. What we do is we target the appropriate R cores for the application. I used to always call it the right core for the chore. You want the best, if it's very deterministic and all that, if you need high performance you select the right core. The last thing I'd say here is the orange box which we introduced last year at Detroit Tech Days. This actually is really interesting because it's like a connectivity domain controller in a way where all the connectivity in the vehicle so if you look this, GNSS for positioning, gyro, digital radio, V2X, even the audio, the radio we have a secure element built into this accelerometer, Wi-Fi 6 and 5G. Anything that's wireless, Bluetooth, low energy, ultra-wideband all of that converged into one box. Now you can do coexistence. There's a lot of opportunities for doing some innovative things with the wireless technologies. This has an i.MX8 XL processor in there. We'll upgrade that over the time with new processors. We're showing here some of the things with Bluetooth, low energy, connection, Wi-Fi. This is a digital radio coming in and we're actually driving the radio from the centralized connectivity domain controller so now you don't have to run all of that to the radio here. It's all done digitally through an ethernet connection. And then we're showing here a Qi wireless power and communication through Bluetooth, low energy also and Wi-Fi with the phone. When I have this box, this is a reference design just to be clear about that because when I first saw it, I said, that's a pretty big box but what this is is a modular, this is more of a development platform where this is an actual reference design. So there's a difference between what's intended to go in cars versus what's more of showing what you can do and this has a lot of modules. So what's really important with this box is that, you know, wireless technology keeps changing over time. So for development purposes, this allows the customers to rapidly move and prototype with different types of modules over time. So it's more in a larger form factor to support modularity and such. That's why it's big. Typically this would be much smaller but these are modules a lot inside right now today. This is a great demo because it shows how everything comes from the wireless through this unit driving the display with information through the Bluetooth and a central point for the wireless connectivity and the Wi-Fi and everything. It's all right there which brings a lot of value. I think you saw the battery management AI and the last thing here, did you see the scooter yet? So the key thing on the message from my perspective on the scooter is that all this great technology that we're developing for the automobile, whether it's the radar, the electrification, the control, the communication, the display board, all that can be resized and repurposed into other types of electric vehicles, right? It's a work with motorcycle vendors, scooter vendors, even like robotic, electric, robotic delivery type vehicles. This technology can really span across a lot of applications. Cool. Thanks a lot. Great to see you again. All right. So there you go. That's a bit of a tour around the NXP booth showing you some of the great technology that we actually have to show to our customers here at the show and thank you very much for watching. Thanks.