 Live from the Mandalay Convention Center in Las Vegas, Nevada, it's The Cube at IBM Insight 2014. Here are your hosts, John Furrier and Dave Vellante. Okay, welcome back everyone. We are here live in Las Vegas. This is The Cube here at IBM Insight in Las Vegas live in the social lounge. I'm John Furrier, co-founder of SiliconANGLE Media, the other co-founder of SiliconANGLE Media, Dave Vellante, and SiliconANGLE News. The Cube, our flagship program. We go out to the events to extract a signal from the noise. Our next guest is Jim Green, CTO of the data and analytics business unit at Cisco. Welcome to The Cube. Thank you. Thanks for coming on. We appreciate it. Internet of Things seems to be the hottest thing going on around these days. All of a sudden, boom, it's on every press coverage, cover stories, trade journalism, New York Times, Wall Street Journal, mainstream media, really picking up on the news. Not a new concept for the folks in the industry. We've been looking at technical papers for going past a decade with data, but certainly now with mobile infrastructure and big data, it's hot. So I got to get your first question on where do you see the current Internet of Things going? Do you think it's crossed over from machines and data centers, or do you think we've now crossed over into the more the mobile personal space? Well, it is absolutely true that the Internet of Things or IoT is the hottest thing going. The number of press articles around IoT has increased tenfold over the past 12 months. So you have to ask yourself, is this a temporary situation? Is it some sort of fad, or is it lasting? And frankly, I believe it's lasting. And the reason that the press has really picked up on it, the reason that the industry has picked up on it, is that unlike some of the other things that we've looked at, over the years in technology, IoT has a solid economic basis behind it. It is quite straightforward to look at a building and to figure out how to reduce its energy consumption by 20%. It is quite straightforward to look at traffic congestion in the city and figure out how to reduce the carbon emissions by 10% or 20%. There is a lot of energy savings that can be taken from the industry and can be taken into effect if you can automate parking. There is a lot of emergency response activities that we can improve by city infrastructure management. We can get more oil out of the ground. We can actually automate our automobiles and make for a higher quality experience in terms of transportation. Lots of really interesting things. So we see certainly the economic side. I had a chance to interview the CEO of GE, Jeffrey M. Elton. They have customers like United Airlines that a 1% increase is billions of dollars. We were at a recent event where, I think it was Tableau or Splunk. I can't remember which one it was. Very small business, but they serve the locomotive, the railroad industry. The savings are in the billions. What looks like small data is astronomical business value. I mean, it's literally off the chart changing how businesses are organizing. Now that's on one end. Now there's also the technology and how they route the data. So talk about that dynamic around the transformative business outcomes and then how that relates to IT infrastructure. Well, there's many different IoT markets. It's going to be segmented by vertical as we've indicated so far. But on a macro scale, you have the consumer-oriented market with the automated home with the wearables, with the Fitbit bracelets, etc. And then you have the enterprise commercial or industrial market. Generally speaking, the researchers are saying that the consumer market will take off first because high volume, low price devices will. However, by about 2017 and 2018, the enterprise or the industrial market will be bigger, both in terms of the number of devices as well as in terms of the dollar amount of the total market. Now you have problems with this enterprise or this commercial business that you don't have in the individual basis and that the amount of data is massive. We're seeing lots of projects where we're going to have over a trillion bytes of data generated every day. And therefore, it actually changes the way computing has to be done. We've done quite a bit of work on this and determined that it actually turns things upside down. For example, in today's world in analytics, you always grab your data in a repository. Out of that population of data, you extract a representative sample and then you do some analysis on it. In IoT, we're going to have so much data that we won't be able to actually store it all. Therefore, we're going to have to analyze it prior to storage. In other words, it's a reversal. It's analyzed before you put it to base. So there's going to be a number of different steps in this process. The devices and their controllers and the sensors generating data, there will be communications and networking to actually communicate the data to where it needs to go. We're going to have to do this edge processing where you can actually filter, reduce, aggregate, and do the initial analysis of data. You then have to store it in some sort of a storage system, perhaps databases, perhaps big data. And then finally, you'll be able to have applications that act on it and analyze it. So as a result of this, the implications to put together an architecture or a model for IoT are much more required in the enterprise space. Recently, IBM and Cisco and others have collaborated to put together a reference model for IoT. And it was presented a couple of weeks ago at the IoT World Forum. And so we now have a model for all of this fits and all these components relate to each other to use as a guide and as a reference as we start to build large-scale systems that do complicated things. Wow, I mean, there's a lot in what you just said. So you're now talking about analysis to data, so that implies real time. And you talked about sampling, so is sampling sort of a dead concept? I wonder if we could start there and then get into it a little bit. Well, I think we need to take the step of analytics and break it down into multiple activities. First, let's take a trivial example. You've got a building management system. The set point is at 71 degrees. You take the temperature that's 70.5. And you turn on the heater. And so, why move that data around? Why not discard that data? Because there's no action to be taken on it. So at the first level, you're going to have to filter what data is important, what data is not important. Secondly, you're going to have to normalize that data so the data coming from a variety of different places can look like it came from one place, so it's easy to work with. Then some data is now going to have to be persisted, put it into a database. Still, even after their data reduction, you're going to have too much data, so you will have sampling as you pull a small set of the data out and then you'll do your analytics using the variety of analytics tools in the market today in much the same way as you do today. But the generation and the treatment of that data with IoT is way different than it is the way things work today. How do you expect to determine in this model this IoT reference model to persist? How is that decision made? Is that automated? Are humans involved in making that decision? Well, there's some people who use the term IOE and while we're still, everybody's got the definitions slightly different. You're a corporate marketing people and you use that term. Yeah, that's right. But in general, IOE encompasses the idea of people as well as inanimate objects and so people in collaboration and it's those people who have to a priori say, I'm a geologist sitting in Houston. We're drilling for oil off the coast of Africa. There's not enough bandwidth to get all of that data back here. So today I want to look at this subset. I want the instrumentation to collect this portion of the data because it's the stage in drilling that where this data is interesting. Therefore, the people will determine what gets stored and what gets communicated and what gets analyzed. So that's some kind of policy that then gets injected and automated throughout the infrastructure. That's right, yes. So a part of an IOT system is that the people have to say, this is what I want to have happen and then the system actually takes care of it. Again, going back to our simple example of building management, you turn to thermostat you're basically changing the policy. But there's so much data. So how can the system how can the reference model, I'm sure you thought this through, help me figure out what my policy should be. I don't even know what questions to ask sometimes. So much data, as you said. So how does that loop get closed? I'm afraid the technology can't tell the people what the policy should be. Oh no. There has to be some sort of human intelligence to actually control the entire system. But there are cases where it's good for the devices to act somewhat independently. For example, if you're in a car wreck, you kind of want your on-star system or whatever system to put in future cars to kind of take over because you may be incapacitated. Similarly, we're going to put telephone chips in pacemakers. And so if you have a bad heart, you fall down you may not be able to dial 911. And so there's a case where the emergency response can be done on your behalf. There's tons of examples. But in general, what we're seeing is a whole new world. You know, we we did the emails and browsing. We did the e-commerce sites. We did the social networking and the Twitter's and the Facebook's. Now, the next generation is IOT, IOE, the intersection of things and people in a much more intelligent fashion than anything that we've ever seen before. You mentioned some of the tension, the pressures at the edge. Historically, the network has been very hierarchical. Everybody talks about traffic moving from north-south to east-west. What's Cisco's perspective on that and how does it relate to the traditional IT network infrastructure? Are they two different worlds? Are there ways in which we can leverage that? Is Cisco sort of accelerating that move to the edge? I wonder if you could talk about the networking implications. Yes, no, we're absolutely accelerating that to the edge. And at the same time we're complementarying that with work at the center. The cloud is going to be a huge aspect of IOT and IOE. And IBM is very interested in the cloud as is Cisco. And so both of us are saying the data center may not be the right place for a lot of this data. We grab the data from the devices. We do the processing at the edge. We move it across the network into the cloud. And the trick is to figure out how to make all of those different aspects harmonious. The big thing for Cisco is that we're obviously excellent at the network. And the network is in the middle of IOT. And so how do we expand upward in the cloud downward into the edge and then horizontally so that we have interoperability across all the devices and all the different vendors. People say, you know, data is the new oil. Data is the new source of competitive advantage. Some people say, no, it's how you apply the data and differentiate that's the new source of competitive advantage. Whatever, if this world is really data driven, data becomes this potential source of value. Who owns the data in IOT? I have absolutely no idea. Everybody wants to own the data. Yes. Well, you know, you're going to establish a system. And that system is going to generate data. And so it's the person who or the organization who sets up the system. There's going to be a lot there already is a lot of interest in city infrastructure management which has to do with parking traffic congestion or emergency response lighting as well as garbage collection. And so the cities will own that data. In a building such as what we said here, the owner of the building will have the data. But then there will be other scenarios that cross various categories where the data will flow for for the assistance of some population and the data will be shared. Give example the pacemaker with the telephone. The individual who owns a pacemaker might own that data, but he wants that data transmitted to the emergency response team. And so there will be lots of gray zones. It is going to be a challenge to auto me and yet retain privacy. And it's another matter that we're actually really studying a lot. So we can't go a minute without bringing in a sports analogy. So I'm going to bring in the World Series because it's a giant span. I'm a Red Sox fan by nature, but 15 years in Palo Alto and I've converted because they get some good vibes. It's very Red Sox-like in the West Coast way. I'm not going to go any further. I could get myself in trouble. You're already in trouble. Not that there's anything wrong with that at all, Dave. But I got to ask you the Kansas City Royals have got great pitching, the Giants are scrappy. So in he choose so's on the stage talking about roles of people. So let's bring the World Series dynamic into the people who really want to make a change and win with big data. That's the practitioners out there who are making changes. They're the ones who see virtualization. They see North-South East-West network packet. They understand virtualization. They understand the mobile infrastructure perimeter with security. These are threshold issues right now in our moment in tech, this tech transformation. So the Royals have good pitching. The Giants get down, but they scrap back and it's about their mindset really. Look at the Giants. This was on talk radio going on, great conversation around their mindset on how they approach every game gives them that edge. So if I'm a practitioner, what approach should I take, given the fact that there's some good pitching to stop me from implementing this new way because a lot of people don't like the change. Especially in large enterprises like change, let's see a POC a lot of debate but a lot of people say let's just move quickly to this new normal of modern infrastructure. What mindset and what approach should that practitioner take? Well I think the mindset we all should take is change is inevitable, but there are so many different types of analytics around professional baseball. It's ridiculous. First off, you know that the guys are in overdrive right now as everybody moves from San Francisco back to Kansas City trying to analyze the strategy for the next game. There's also the people who actually own the stadium and the franchise. How do you move people through the turn style quickly? How do you organize parking? How do you handle the traffic once the game is over? Tons of analytics in order to actually improve the quality of the experience. We're seeing analytics applied with the replay which we now have in professional baseball, a whole new scenario. And also if you think about it, the excitement of the game is to be in the moment but you leave one thing behind when you leave home and that is the instant replay. Yet we all have video devices that we carry with us and so we're now seeing data moving off the field into the studio into Verizon and your carriers onto your phone so that you can actually have the experience of watching it live and immediately seeing the replay. But then on a more commercialized basis a lot of revenue comes from the concessions, the food stands, et cetera, and the stadium. Dynamic pricing, based on wins and losses, that sentiment. Essentially baseball is interesting because if you look at how progressive they are, they have a really hard challenge. They have to manage their club. They have to manage their organization which is the employees and all the dynamic pricing, profit and loss and the fan experience. That is what enterprises need to do. They have fans, that's fan experience called customers. The net of all of this is that baseball is 100 years old and yet it gets more exciting to go to a baseball game every year. Because of all of this combined it enriches the personal experience. Lessons for that. Let's extract some signal out of that analogy because it's really relevant. Certainly they are under scrutiny. Baseball is the crowd sourcing aspect of the truth and the crowds. Everyone is watching, the speculation. They are harnessing the data from the crowd. They are harnessing the data from all their databases. What is the lesson for IT? Just speed. Building a dynamic system. For example, the cash registers are telling the computers throughout the course of the game how darts are being sold. It's absolutely true that real-time advertising can trigger increased sales. At the seventh inning stretch, you need to know what to put up on a billboard. Very real-time, very dynamic in order to actually get the maximum revenue out of the experience and also to enable the fans to actually know what's out there, what can I do to make my day better. This is Beyond Moneyball. We should do a segment on this with Cisco in particular because Jim McHugh, Cube alumni, manages the MLB relationship. It might not be known out there, but Cisco provides MLB with all the gear for their stadiums and all the different teams. I know you guys are close to it, but this is Beyond Moneyball because that convergence, the confluence between team, organization, fan experience is an overlay for what an enterprise could do to be rolling stuff out. It's about the business owners inside the company to deliver on all three of those objectives. Every major league stadium has miles of networking in the stadium. Lots of networking devices. Again, advertisement, big displays, concessions, lots going on electronically around the baseball game. Jim, I got to ask you relationship with IBM, you guys with UCS, server business, IBM just sold its x86 business, so that opens up a whole potential set of collaborations between IBM and Cisco. Coincidence, cause and effect, what's the relationship with IBM today and going forward? Well, Cisco and IBM are major partners and have been for a number of years. Each of us sells a ton of stuff to the other and buys a ton of stuff from the other. There's strong economic ties for us to figure out how to go into the market to coexist so that we can provide a better computing experience for our combined customers. Now, from time to time we make some moves which make things better or make things worse. IBM selling off their small computer actually opens up opportunity for both of us to base more of our systems on UCS. But IoT also is an opportunity. IoT is an early market and so in the course of time I've come to know the IBM folks and work with the IBM folks and as I said before, we're working together with standards bodies and working together with industry consortiums. We're on stage with each other as happening today and happened a couple weeks ago and so we're trying to promote the industry. We're trying to get the industry to actually move forward and leverage what is possible in a way that benefits all of us. Right, and others potentially in the ecosystem like I said, it's early days. It's a very complicated ecosystem and one that we're following, John. Okay, Jim, really appreciate you coming on. I mean, I get to talk about MLB and big data because I really think that what's going on in sports really is a predictive aspect and we can bring some cognitive analysis to that conversation and have fun with it. Thanks, Gaston. We're live here in Las Vegas and IBM's doing something really special this year with social VIPs. They're handing out these awesome badges and what they're doing is really bringing the crowd into the conversation. They're bringing in crowd chats. There's a crowd chat at 12 o'clock, crowdchat.net slash IBM Insight with Brian Fonzo and Carla, Data Nerd and it's a thought leader. It's organic. It's an unconference within the conference. It's a special digital experience called Insight Go. We're glad to be a part of it and the VIP social influences are bringing a whole new level of experience here at IBM Insight, awesome great, great crew here. So we're happy to be a part of it. We'll be right back with more analysis and conversations here on theCUBE after this short break.