 Live from San Diego, California. It's theCUBE, covering Cisco Live US 2019. Brought to you by Cisco and its ecosystem partners. Welcome back to San Diego, everybody. This is Dave Vellante in Stu Miniman, and you're watching theCUBE, the leader in live tech coverage. Lisa Martin is also here. Stu, I actually did see Ron Burgundy on the street last night. He was out, he was shaking hands with all the CCIEs. This is day three of Cisco Live 2019, theCUBE's coverage. Joe Malenfonte is here, he's the director of IoT marketing at Cisco. Joe, great to see you in from Colorado Springs. Thank you very much. First time in theCUBE, welcome. It is my first time in theCUBE. Thankfully it's not actually just a box because I have a little claustrophobia going on. So, IoT, it's got all the momentum. Liz Santoni was up on stage this week addressing 28,000 plus people. What's driving all this momentum other than the great marketing? What's really happening in the field? IoT has been a very nebulous thing for the last few years and we're finally starting to see some solidification and convergence around what it means. And really for Cisco, we started on this path a few years ago but Liz took over last year, we've established a new strategy because customers, organizations, and especially organizations that run operational technology. Think of refineries in the oil and gas industry, in the electric utilities industry. They run a whole separate network called industrial control systems and that OT side of the house has traditionally been very siloed. Well, as the economy moves forward, as we digitize, they're trying to connect back with their enterprise side of the house. Well, if you're going to connect your network with the IT side, why not use the incumbent leader in enterprise networking? We know who they are. We're all sitting here right now with Cisco. So they look back to the IT side to say, hey, please help us connect. That's really what's driving the market. How should we think about the difference between OT networks and IT networks? Are there any things we can learn from Telco, which also had some unique attributes to it? But share with us what you guys have learned there. So the OT network is very different, right? It's very time sensitive. Latency is just something that they can't have. When you think of email going down, what's the worst thing that happens, right? You might get a nasty gram eventually. Well, when the power grid falls over, lives are at stake. So those networks are very critical. They're very sensitive, and they've always been kept separate. So as they start to make that interconnection, we need to bring together networking technologies that are for that environment. As they make that connection though, there is a very number one concern for them is wait a second, if we're going to connect this stuff, we need to make sure it's secure. If you're a chemical processor, for example, and you've got a secret recipe, you don't publish that. You don't patent it because you don't want the word to get out or else somebody's going to rip you off. So they don't want to have this side of the house get connected with that side of the house and expose the secret sauce. So security becomes very top of mind for them. Connect it, but do it securely. All right, so Joe, I've actually been happy with how I've seen the solutions maturing from Cisco because when I first heard IoT, it was like, well, we're the leader in networking, we're going to network everything. And I'm like, okay, but at the device edge, one of the challenges is, often I have limited or no connectivity. So sometimes I'm going to need to do the processing there. There's lots of different protocol issues that I have there. So talk about some of those new solutions that Cisco's been doing at the edge that are more than just sending bits back and forth. That's a great question too. So of course everything has to do with networking, but networking is merely the vehicle for connectivity. And so we realized very quickly is if we're going to create new routers and switches for this environment, there's an opportunity to do a little bit more. So back in February, we did something at DevNet Create called a hackathon. We have a new router, it's a ruggedized router called the IR1101, but I think Liz showed it on stage the other day. And this has a specific module inside of it. So there's a module that can be swapped out. Well, at the DevNet Create hackathon, one of the teams actually created a machine learning module. Why machine learning at the edge, right? If you have 700 substations, you don't want to deploy machine learning on each and every one of them. You want to get all that data back into a central place so you have more data to actually train your algorithms on. Why would you put ML at the edge? Because not everything needs to come back. There's stuff that you can do at the edge, number one, with that machine learning on traffic that doesn't have to go back. When you don't backhaul traffic, that means you don't have to pay costs over to your LTE carrier for more data. Other times as well though, they're looking at compliance as another reason. So in terms, that's one use case, right? Let's look at the other one, which really comes down to, okay, if I'm connecting things and you can actually do some computing at the edge, how are we going to do it? On all of our new switches and routers that have edge compute capability, they're using native Docker containers. So now you can actually deploy your applications at the edge. Again, do the work at the edge as close to where it has to be as possible. Don't bring it back. You don't have to worry about any sort of violation of compliance with local laws, sovereignty clouds. You don't have to worry about costs of backhauling traffic. And then if anything's time sensitive, it stays as close to the edge as possible. All right, so one of the keys here to your strategy, it's clear, is to allow developers to build new applications at the edge. You're not OT experts, that's not your roots. And those developers at your ultimate clients are. They're the domain experts. They know what's going on. They know these specialized areas. So talk about the importance of having programmable infrastructure at the edge. And specifically what your strategy is, where does Cisco leave off? And where does, you're not a pass vendor. You're going to bring that in through partnerships, but help us understand that strategy a little bit better. Our ecosystem is incredibly important to us, right? So we've got, DevNet is incredibly important to Cisco because as you heard probably on yesterday, Susie announced new certifications for IoT. Those certifications allow engineers, whether it's a control systems engineer, whether it's a network engineer, to actually get certified, be a specialist, be a professional in writing their applications for the edge, for those specific environments. But more importantly, because let's go back to the environment that we're working, right? Time sensitive, very critical, low latency networks. You don't want to go and push out something where you're not 100% certain. So IoT certifications that DevNet has created give those engineers a repository, a sandbox, and all of the Cisco solutions to actually test with before they do the deployment and ensure, almost guarantee themselves, success by pushing to production. Yeah, and one of the key things there, Stu, is the ability to test things quickly and fail fast. Yeah, well, one of the things I was a little bit concerned about when we saw this wave of IoT is, every customer's going to have different requirements. So it feels like we at least get some level of maturity and commonality if we can have certification. Exactly. What does somebody come out of? What skill set do they have? And right, because you said as if I'm a manufacturing or healthcare, everybody's going to use IoT, but how we're going to use it, where we're going to use it's going to be very different. So what's that base layer that we're learning about? So ultimately, the engineer who's actually coding these things, kind of what you said, right? They're all going to be very vertical specific use cases. There's not a lot of horizontal stuff going on. So we're creating a baseline for the engineer to understand their environments better. They honestly know it better than we do, but we want to make sure that as they go deploy these things that we give them the infrastructure to do it on, the application and framework within which to do it, and the tools to be able to do it. And so that's the docker, the modules of being able to do edge compute, and then lastly having that certification within IoT to how do I code this thing? Can I guarantee that I'm going to be successful and push it out? Joe, what's the organizational dynamic like? You always hear the stories, OT is not talking to IT. They're different animals. You got some hardcore engineers that have hardened their infrastructure, and you got IT guys that are trying to build applications and support applications for the business. Those two constituencies don't talk. What can Cisco do? What's the strategy with regard to bringing those constituents together? Do you have to? Or is it sort of divide and conquer? I think the number one thing that we want to do is enable the collaboration between the IT and OT. It's not that people don't want to. They're just trying to figure out how to do it better. So if we can help them, number one, connect their networks together safely and securely, that's number one, reliable and secure networking, what we're known for. But number two, from the OT side, back to what I said originally was around the security side. So I don't know if you guys heard, we announced last week our intent to acquire a company called Centrio. Now why is this important? Because they do passive network detection, whether it's anomaly detection, but they do asset discovery as well. Now a big thing when you're connecting those OT networks into the IT world is what assets do I even have? Those assets are vastly different from anything IT actually knows. So this acquisition will allow us to passively discover and tell them, here's your list of assets that you're going to be connecting, here's what we need to secure, so they know in scope as they walk into this project, they've got a really good blueprint for what needs to be done and not surprises. And the reason that's important is about only 40% of all IoT projects make it from pilot to production. I mean that's kind of staggeringly low. I actually had an analyst tell me yesterday, I'm shocked you guys said 40% because I only hear about 30%. And when you're doing it in a lab, you know all the variables, but when you go out to a brownfield environment where you've got 20 year old systems that honestly was probably a system hidden underneath some guy's desk that nobody's actually known about, if it ain't broke don't fix it. We're actually able to discover all those assets now. That's why we did the acquisition. So it's really from an asset visibility and a security standpoint. And you're saying Joe that those, that discovery is specific to edge assets versus like a stealth watch. We heard a lot about stealth watch this week which is, they do discovery, but you're saying that's predominantly for IT assets. Servers, storage, networking, searches, et cetera. Yeah, I mean listen, so stealth watch is awesome and I think eventually there's going to be a little bit of a merger between some of these things, but the operational technology environment is very different. They don't, they're not native IP. They don't talk the same protocols. There's thousands of different protocols that exist in an operational technology environment. DNP3, Modbus, ProfiNet, Profibus, just very few right there. Those, IT has never ever talked them. They don't even know what they mean for the most part. Tell an IT guy, hey can you detect this DNP3 traffic? The answer is no. However, when we move into that environment, our networks need to be able to understand that traffic and that's where Centrio comes in with that operational technology expertise to help the IT and the OT really come together. Business outcomes at the end of the day. So Joe, give us a little spin around the show from an IOP standpoint. We've got the IOT takeover happening here in the DevNet zone. All the classrooms seem pack jammed as they've been all week for all the takeovers but give us a little spin around. It's been amazing actually so far this year. Having been at Cisco for a few years now, I walked into this and say, wow, we are definitely in the IOT world. We've got IOT plastered outside, we've got it inside. People are very interested in IOT. They're not just in what we're doing but how they can take the knowledge and what they're going to learn here and really bring it back into a practical use case at their own organizations. So from an IOT perspective, the world of solutions downstairs is jam packed. I mean, we've got a massive presence down there. We've connected the buses that are outside. We've given, if you look at the app, we've actually connected those buses to the app for real-time data to say, this is when the next bus is actually coming. I mean, what a pain in the butt is it to stand outside and go, where's the shuttle bus? We can tell you where the shuttle bus is. We can tell you when it's coming and how long you're going to have to wait. And yes, don't worry, you've got time to go get another coffee. Yeah, just follow the line. You'll find the bus. You'll find the bus, but how long is it going to take to get you there? Okay, you were mentioning about some of the reasons for apps at the edge. I want to come back and explore that a little bit. You said compliance, I think it's true in cost. There's physics involved as well. So the cloud guys would say, hey, we know there's a lot of stuff going to the edge, but ultimately the heavy work is going to be done in the cloud and all the modeling. You've got others who're saying, hey, here's a box, we're going to put it at the edge. You know, it's kind of a top-down approach. Why, what's your scenario in terms of data at the edge? Why does data need to stay at the edge? You mentioned real-time before, but let's double-click on that a little bit. So I think there's really three key reasons that data and applications are going to be processed at the edge. Number one, compliance, right? So there's certain data that's going to come in that cannot be shipped back to a public cloud. That's part of the rules. You cannot do it. No public cloud storage for certain private data. Number two is cost, honestly. And this is a really big one. If you can reduce your overall cost instead of backhauling all that traffic to HQ, to your data center, and you just keep it at the edge, you don't have to backhaul it. LTE traffic, not the cheapest, and I can only imagine with 5G how much that's going to increase the cost. They're going to want to just backhaul everything, right? We can do that really quickly. We can take everything and put it back. Yes, but your bill every month is going to be monumentally more expensive. And then lastly, as you mentioned, was the time-sensitive one. That's really going to be one of the bigger ones from a business standpoint. The engineers are now going to be able to write applications for processing data at the edge so that they don't lose, in this environment, three seconds is the difference between life and death. I'm kind of exaggerating, but kind of not. If you're missing an alert in a couple seconds where you can't shut down a valve, a gas leak valve, where there's potential for an explosion, those seconds are the difference between boom or we're all good guys, it was just an alert. The other classic example you hear is autonomous vehicles as well. You can't run that from the cloud, right? You've got to do that locally. Last question, Joe, is Cisco differentiation? Obviously you come at it from a position of networking strength. You mentioned that in your opening comments, but give us the bumper sticker on why Cisco. I think that the big reason why Cisco is unique in the IoT world is, number one, we're not trying to be everything to everybody, right? We're trying to create a safe and secure and reliable network. Number two, though, is our ecosystem. So we have a large partner ecosystem. We're expanding it into the OT world. We've got specific products for those OT partners where they can embed our networking technology into their solutions and systems that they're putting together. Lastly is honestly what we're doing here with DevNet. Nobody in this world other than Cisco has DevNet with the network, with the ecosystem. When you put that trifecta together, it's unstoppable. And so being able to bring together IT and OT, only we can do that with those three things. So yeah, I think Susie said yesterday, Stu, 600,000 engineers sort of trained on coding, Cisco infrastructure. It's going to be interesting to see how the OT folks pick up on that and what the adoption is there. Joe Malinfoot, thanks so much for coming on theCUBE. It was great to have you. Thank you, gentlemen, I appreciate it. Really a pleasure. Okay, Stu and I will be right back. Lisa Martin is also in the house. You're watching theCUBE. We're live from Cisco Live in San Diego. We'll be right back.