 Live from Mountain View, California, it's theCUBE, covering DevNet Create 2019, brought to you by Cisco. Welcome back to theCUBE, Lisa Martin with Cisco, DevNet Create 2019 at the Computer History Museum, been here all day, talking with some really great innovative folks, excited to welcome to theCUBE, Marty Jane, Senior Director of the Cisco Global Partnership at NVIDIA. Marty, it's great to have you here. Thank you, good to be here. So I always love talking about partnerships, we're at day one of DevNet, tomorrow's day two, there's been a lot of community spirit is here. So I just kind of in the spirit of partnerships, a lot of collaboration, that community is really strong. Before we get into kind of the details of the Cisco NVIDIA partnership, first kind of thing I wonder is, all right, this is the developer community. Why the developer community with NVIDIA? That's a great question. So if you think about, we make GPUs, which is a piece of silicon, a graphics processing unit, and it is really only a piece of silicon until a developer comes along and develops a cool app on it. So if you think about how we go to market our large conferences called GTC, it's really developer focused. We have a little over a million developers in our ecosystem, and I find it very synergistic with Cisco, if you think about Susie Wee's vision, I think it's the same idea. You look at over half a million developers in their ecosystem, and they want to develop cool apps, and that's how your platform becomes relevant. So if you think about all the modern innovation, that's coming from developers. So these are the folks that we should be talking to on a daily basis. So I see a lot of commonality, a lot of synergies. In fact, we had Cisco DevNet come over to our conference GTC, and they appeal to our developers, and now we are here talking to their developers, and we're also developing some joint platforms which these folks can use for, like I said, the more modern apps with all the new data that is coming whether it's the IoT, whether it's AI machine learning, automotive, smart cities, you name it, we need to be able to provide that platform to the developers. And a number of those topics came up today even during the keynote, smart cities, being able to utilize and accelerate workloads with AI and machine learning. They gave some great examples during the keynote of how developers can build networks. They give this cool example of, I think it, right off the heels of Coachella, of designing a secure network for an indoor concert, designing it for an outdoor festival like a Coachella, and then designing it for a massive stadium like a big football game, like the Super Bowl, for example. And they showed at that higher end, they showed how they're using machine learning to zoom in on, for example, they had this little red box and you see people. And what's actually in there that the machine's detected was a fight. And in real time analyzing this data and then dispatching the appropriate security to come and obviously probably take the drinks out of their hands first, but it was a really interesting, great real world example. So you guys have been partners a long time, or you've been actually working at various companies with Cisco for a long time, but I think of Cisco and NVIDIA coming together. How are you guys helping to accelerate these AI and machine learning workloads that we're starting to see in every industry? You bet, great question. So let me first comment on what you said about smart cities. I like to think of it as smart and safe cities. So actually the first set of applications will be around public safety. What the example you were giving is spot on if you have large crowds gathering, it makes sense for us to be able to look at those crowds, we call it intelligent video analytics or IVA. In fact, we have a platform here, the Cisco IR1101 with a GPU added to it. So now I can watch the crowds and if there's a fight breaking out or somebody's carrying in a weapon, you want to know. Somebody walks in carrying a backpack and drops it and moves on. You want to know and inform somebody. So what is happening is we have these millions and millions of bytes of video data and that data is not being really used today. So what we're doing is saying, you know what? Let's find those pieces of intelligence in the video data and do something with it. And public safety is absolutely the highest priority so smart and safe city makes a lot of sense. So what we're doing is we're going to market with partners at Cisco. So what we're doing is we're saying, okay, let's design these GPUs into these servers which are connected to cameras and think about how many cameras are deployed today? Probably a billion and a lot of the video data can now be used for public safety purposes and we basically go out and talk to large companies, we talk to governments, we talk to cities along with Cisco to go even open their eyes to what is possible today. Right, because if that data is dark for so long they don't know what they don't know. Well, in most cases what happens is you record four days of video and until something happens, nobody goes back and takes a look at it. But now we have the ability to look at it real time and cities and governments desire that very much so. So an example, it's such a relevant topic, I mean, you know, there's also the issue of privacy but to your point about not just a smart city but a smart safe city, I like that. I think it's absolutely imperative. How do you have those conversations with cities, with governments about, all right, this is what we want to do. We want to actually apply machine learning so that the machines are taught what that line is with privacy, what those boundaries are so that a person, I'd say a lay person, not in technology, maybe as a city government official who doesn't understand the technology or need to, will go, ah, I get it. Yeah, so our conversations are really about what we call use cases. So think of enterprise, a good use case would be, in fact, we work with Cisco on developing this use case. You know, you always badge it into an enterprise. You have your badge, you walk in but you also have some cases people following you in. What stops you from following me into a building? And usually people are too polite to say, no, you can't walk in behind me. Because we've all had the video training or read the manual, we know we're not supposed to do it. We're not supposed to, but then you're like, I don't want to be rude. We're just, culturally, exactly, we just can't say no. Not that I ever heard that. So now we have the ability, so we trained an AI network to say, look, if Marty's badging in, only he's allowed to walk in and if there's a second person who walks in, I want to take, put a little red square on that face and inform security that we have had more than one person walk in. So these are some of the ways, so we talk about use cases. This is one use case. Crowd behavior analytics is another use case. You know, people are walking with a backpack, dropping it. Other use case would be something like BART. So BART loses millions of dollars a year because people jump the turnstiles. And BART didn't really have a good way of monitoring, measuring the losses until we put a camera and captured the number of people that were jumping the turnstiles or going in through the handicapped access. They were losing 10 times the dollar value of what they had thought. So this is how we started the conversation with use cases and what would you like to do? Being able to count the number of cars in an intersection, being able to count a number of pedestrians so you can do traffic management better. That's the language we would use with cities and governments. And then we go deeper as we go through the implementation process. Well that makes perfect sense going in the use case route because you can clearly see in that example that you mentioned with BART, a massive business outcome and an opportunity to regain a tremendous amount of resources that they could redeploy for whether it's new trains, new trucks, et cetera, them not realizing we're losing how much money? I think anybody, when you can put the use case in that context of this is what you can expect as an outcome, they get it. Absolutely, that's really the only way to start the conversation than starting from bits and bytes. And this is usually the case across industries. So if you think about retail for example, you go to a Safeway and you start talking about GPUs and servers, that's not the great way to start but they do have issues with shoplifting for example. So how do you know a person is walking in through the checkout and they have one item and there's a smaller item right here and they walk out with it. So how do you monitor that? So now you can do that with the right kind of cameras that can capture, look there are two items, not one. How do you know where a shop are stopping? Which aisle is the most popular aisle? How do you know that? Well now you can have cameras that say look, we have red zones and green zones so you can do those kinds of things with modern ways of doing AI. So interesting because it's so, I mean the examples that you gave are so disparate but yet they make so much sense is how you're describing it rather than going into a grocery store and talking about GPUs which they might fall over with their eyes doing this, you're actually putting in the context of a real world problem they've been experiencing since the beginning of time. And they understand oh my goodness, this is how we can use technology so Safeway becomes a technology company and they don't know it to actually start impacting their bottom line. That's right and so now if you take that and you extend that into how do you go to market and this is something you wanted to touch on how do you go to market with Cisco, how does NVIDIA and Cisco do it together, right? So think of Cisco's sales teams who are talking to all these customers every day whether they're retailers, financial services, federal government, healthcare, you name it. So what we've done is we basically sort of take in all these industries and created the top three or four use cases we know are relevant to that industry either for safety or for saving monies or for a variety of their operational reasons and we have narrowed it down to three or four or five use cases in each of those target industries. So what we do now with Cisco teams is we would bring them into our facility or go to them and really talk through all those use cases and train them on hey look, this is what we do jointly and that makes the conversation much easier then they will go and present to the customer and once the customer gets an idea of this is all possible now that starts a deeper level of technology and server and GPU engagement. So this is one way we go out and talk to different customers with Cisco today. So let's stick into that a bit more because Cisco is so enormous. Yes. They have a billion different, I'm slightly exaggerating, products, but a lot of different technologies that form many different solutions. So I imagine your Cisco expertise over many years of working with Cisco as a partner for other companies. How do you, once you get to that deeper level conversation, how do you bring those different groups within Cisco together so that that solution conversation is one that really aligns to that use case and the customer goes, oh I get it. Yeah, that's a difficult question to answer. That's like, you know, Cisco is a large company but I think I also think they're also very sales driven and that's what drives the different groups to come together. In fact, some people called me the connector because I've been working with Cisco for so long. I know people in DevNet, I know people in sales, I know people in the server BU. In fact, if you think about the platform I was talking about the IR1101 with the Jetson GPU that came as a result of us talking to the IoTBU, result talking to DevNet, Ashutosh and the DevNet, he said, you know what? This is cool, I'll go do this. Then we took that to the IoT guys, he says, oh this is cool, we can take that, put it in this platform. And then, actually next week I'm talking to a sales team at Cisco, they cover utilities and this platform is perfect for utilities. Even think about like fire monitoring in a forest. How do you deploy thousands of people to just watch what happens? We can take a platform like that now and really deploy it in hundreds of places which can monitor fires or the starting of a fire. But yes, bringing them together is no easy task, it's fun. Well you are smiling, I like that. Well Marty, the connector, Jane, thank you so much for joining me on theCUBE this afternoon, fun conversation, I enjoyed it. Of course, thank you likewise, thank you so much. Thank you. Lisa Martin for theCUBE, you're watching us live from Cisco DevNet Create 2019. This is the end of day one, stick around, John Furrier and I will be back tomorrow to cover day two, thanks for watching.