Introduction
This week I’m joined by Mark Thiele, the CEO and Founder of Edgevana.
In our conversation, we dig into what the edge to cloud landscape looks like. We talk about how enterprises will use each of these methods along the spectrum of edge to cloud. Mark outlines why the spectrum matters for enterprises and details examples for each.
Speaker Profiles
Mark Thiele Twitter: https://twitter.com/mthiele10
Mark Thiele LinkedIn: https://www.linkedin.com/in/markthiele/
Edgevana: https://www.edgevana.com
Audio File
Featured Hashtags
#CIO #Edge #IoT #Cloud #DigitalTransformation #CxOitk
Podcast Transcript
Tim Crawford 0:00
Companies are looking for new ways to transform their business. Technology plays a critical role in this transformation. Speed and innovation in both technology and thinking are key to this shift. Hello, and welcome to the Cxo in the Know podcast, where I take a provocative but pragmatic look at the intersection of business and technology through the lens of leading CxO executives. I’m your host Tim Crawford, a CIO and strategic advisor at Avoa. This week, I’m joined by Mark Thiele, the CEO and founder of Edgevana. In our conversation, we dig into what the edge-to-cloud landscape looks like. We talk about how enterprises will use each of these methods along the spectrum from edge to cloud. Mark outlines why the spectrum matters for enterprises and details examples for each. mr. Mark Thiele, how are you today?
Mark Thiele 0:54
Hey Tim, I’m good, man. How are you?
Tim Crawford 0:56
I’m doing great. So, Mark, just to remind our audience, CEO and founder of Edgevana, and today we’re going to talk about mapping the landscape, looking at edge to cloud. So I think this will be a really interesting conversation.
Mark Thiele 1:11
I’m looking forward to it. There’s a lot to unpack there.
Tim Crawford 1:14
Well, we only have a short period of time, so let’s get right into it. So, what does this landscape look like when you think about edge to cloud, what does it look like, and what’s different from before?
Mark Thiele 1:26
I think there’s a number of things that are different, and and some that will stay the same. And and staying the same doesn’t necessarily mean bad or good. You know, cloud will continue to support and offer many of the services that it historically has, and will extend some of those services via platforms like the Wavelength Agreement between AWS and AT and T, or putting cloud stacks in people’s data centers and calling them edge, like Azure Stack or Anthos, etc. But there are a lot of areas of opportunity where cloud, just by virtue, can’t be and isn’t distributed in enough places close enough to where the work is occurring to satisfy what is likely to be the demand when edge really starts to get going, and some companies are seeing that already. And so, you know, the landscape will continue to have public cloud. It will continue to have private data centers and colocation data centers, but specifically between public cloud and edge, edge is a new beast. And while edge may benefit and or be able to leverage public cloud on occasion, and even maybe vice versa, they are not by necessity connected to each other.
Tim Crawford 2:40
So edge is not cloud per se, but there might be relationships between the two.
Mark Thiele 2:47
Yeah, I mean, in theory, some people are working on these technologies to allow for something that would be considered a cloud in the sense that it’s at the edge, distributed in a lot of places, and those places all look and feel as if they are one place relative to the ability to share workloads and failover and things like that.
Mark Thiele 3:04
Sure,
Mark Thiele 3:05
but by virtue of being high capacity, high bandwidth, always on locations that can crunch and solve for any problem that a business might have, edge is likely going to stay fairly different from what public cloud can offer today.
Tim Crawford 3:19
Why do I care about this spectrum? You know, I’ve got my corporate data center. I might be doing some things in cloud, just at a top line. I know we’re going to get into some of the details about this and pick it apart, but just at a top line, why is this spectrum of different modes? Why is that important? Why should I care?
Mark Thiele 3:37
Yeah, that’s a good question. I mean, and for a lot of people, you may not care right now, and I’m going to start with a. I love analogies. I’m going to go ahead and say I’m self-proclaimed king of analogies, which is probably wrong. I love analogies, and to take an analogy I’ve used in the past and say, Tim, let’s say that you’re a newlywed, and you and your wife are going to go buy a car, but you’ve talked about having kids, but you don’t have any yet, and you’re going to go buy a car, and so you look at the cars, and you’re thinking, “Well, I’d really love to get that two seater Miata, but if we’re thinking about having kids, maybe I should consider something like a Camry instead. Or if I’m going to buy a Camry, even though today I don’t really care about the sunroof because of where I live, I might be moving, and so having a sunroof might be interesting, and/or it might help in resale, et cetera, et cetera. So those are options and concerns that you make based on how you recognize your future may change and your future usage characteristics may change. And so the way people are, I believe, the way more and more people in the IT and technology space are beginning to look at technology utilization and deployment strategies are around what you and I have talked about a couple of times already. You know the idea of best fit IT, but also around the idea of being able to position themselves to be able to leverage technologies and services more effectively and more efficiently. And to steal a term that’s common in the industry. With more agility, and so something like Edge is, by virtue, an opportunity to review and to look at and to see how it may fit in business model opportunities going forward.
Tim Crawford 5:12
Okay, got it. So I’ve got Edge. I’ve got my corporate data center. I might be using some colocation. Might have some private cloud in there, public cloud might be using some SaaS. Doesn’t this make it more complicated though than to manage and therefore potentially increasing my my risk footprint?
Mark Thiele 5:31
No, it it can. You know, not only from a risk but also from a you know true cost of ownership standpoint. The simple truth would be it’d be great if we could still run our entire business on one 386 PC with a 40 megabit hard drive, and some people did in the late 80s, early 90s, ran a whole business on a 386 PC or a 286 PC at the time. It was late 80s, but what most people recognize is that at some point, at least the more forward thinking organizations, at some point, complexity can’t be the reason to ignore an opportunity, but it is a reason to tread carefully and ensure that you’ve thought through the business opportunities effectively. But for most, at least of the people that I talk to, and certainly from my own personal perspective, where edge provides opportunity, it is a major step forward, an opportunity that isn’t just about I’m going to spend another five or 10% on my infrastructure costs or overhead, or have to come up with some new threat and compliance strategies in order to mitigate those risks, but rather that this creates an opportunity for me to differentiate myself from my competitors with my customers, and better differentiate my product strategy, or better deepen my relationship with customers or partners, etc. And therefore, the value is there, right? I mean, it’s a topic I’m certain you and I have talked about in the past without ever talking about edge. Is that IT shouldn’t be looked at as how much does it cost to solve this problem, and if it costs more than the problem, then you shouldn’t do it. I should be IT should be looked at while that may be a component. IT should be looked at from the perspective of what kind of opportunities do IT spend create, and are those opportunities potentially greater than the cost or the risk associated with the activity in question.
Tim Crawford 7:24
That’s right.
Mark Thiele 7:25
And for many, I think they’re going to find that for the edge, that’s true. That they will find that there is opportunity there that justify that extra effort. And certainly, there are enormous numbers of companies out there helping to try to find ways for you to mitigate that risk and compliance concerns and security threats and management overhead, many of those companies that are fantastic in what they offer that didn’t even exist three or four years ago.
Tim Crawford 7:52
And you bring up a really good point about how to think about those costs and complexity. I’ve heard something very similar in terms of costs around cloud, for example, and you get the bill from cloud, and and then you freak out, and then you move away from cloud, and that’s where terms like repatriation start to come into play.
Mark Thiele 8:10
Yep.
Tim Crawford 8:11
But I completely agree with you. I think it’s important not to say, oh my gosh, we’re spending a million dollars a month on cloud services or edge services or having to manage this complexity, but rather to say, okay, what are we getting for that million dollars a month? And if you’re getting $5 million a month in revenue, then I think the question becomes, okay, instead of spending a million dollars per month, how do we crank that up to $100 million a month, and therefore potentially ending up with 500 million in revenue. It’s about value, right?
Mark Thiele 8:44
Yeah, absolutely. If you don’t mind me dropping in a few more points relative to that, really quickly,
Tim Crawford 8:49
please do. Is
Mark Thiele 8:49
you know you think about how IT is, and I’ve written obviously about this considerably and contributed to other people’s work even in this thinking. But you have to be really careful. In my mind. You have to be really careful about making assumptions of the value of any one thing that you acquire. So, I’ve got an ax in the garage. Stupid example, but I’ve got an ax in the garage. If I needed a wedge to keep from somebody opening my door, I could use that ax and put the blade underneath the edge of the door, and it would become a wedge instantaneously. Did I buy it to turn it into a wedge? No. But did I just increase the value potential for that ax? Absolutely. If I bought a car and don’t drive it very often, could I potentially get value out of it by turning it into the car that takes all of my neighbors’ kids to school at the same time in order to get more value for everyone out of the car. Maybe I didn’t buy it thinking I could do that, but absolutely. So, if you buy something thinking about only how you’re going to squeeze it rather than how you’re going to leverage it, you’re creating a mindset. And I would caution folks that it’s great to. Be efficient. I have had an efficient bent in my career my entire life, but it’s really dangerous to set metrics that only pertain to things like how much can I shave off of this, rather than how can I leverage this. And by virtue, to your point, Tim, by virtue leveraging it more and actually making money out of the investment that otherwise wouldn’t have have existed, and potentially deciding to invest even further.
Tim Crawford 10:29
And I think those are great points. We’re talking about edge. We’re talking about cloud. Let’s talk about the relationship between the two. And I’ve heard this come up a couple of times just this week in in different conversations about the relationship between edge and IoT and cloud, are they the same? Is one required for the other? Let’s try and break that down a little bit and help folks understand how those two work together or don’t work together, are the same or aren’t the same.
Mark Thiele 10:57
Yeah. So let’s start at the small and work towards big. Right. So let’s start at IoT, and IoT. First of all, IoT is not just IoT, right? I mean, a car could be considered an IoT device, or it could be considered 1000 IoT devices. An RFID tag, in theory, or something similar to an RFID tag, could be considered an IoT device, as could an 8K camera that does facial recognition or just records the environment around it as a sensor. So IoT first and foremost is not just any one little device or pinhead type object somewhere. It can be almost anything, including bits and pieces inside of machines on a factory floor, or something hanging from a telephone pole, or as we already spoke about something that’s riding along in your car, and IoT in many cases, not by necessity in every location, but in many locations, IoT is by virtue edge. But the part of the edge story that becomes important relative to IoT is why are you using IoT in the first place? That’s one. And second, how much data are you creating, and what’s the importance of that data, both in security for your company, IP or other, and in value from a perspective of how it will be used, and is that value derived from time, or is it derived from quantity and quality, or some combination of the above? So all of those decisions go into helping you decide whether I need edge or whether I can do something in in my on-premises data center, even though I have diverse locations, or if I need to do something in public cloud or all of the above. But generically speaking, when people talk about IoT and they talk about edge computing, they’re talking about the assumption that in many cases the data being collected at the edge will have more value the faster you can react and respond to that data, and because data is heavy and data creates latency when shipped to places, those are things to be avoided if you want instant access to the information that data, in theory, holds. And so, the notion that you could just capture video feeds from 100 cameras on a stadium that are running at 4k and just ship that all to some data center 1000 miles away, assuming you had that kind of bandwidth to begin with, but then get a result from it and get it back to you in a reasonable amount of time, especially if that time was related to a a live event occurring in that stadium, frankly, that’s just a ridiculous assumption. It’s just it’s just not going to happen. And when you look at the proliferation of devices, whether it’s more use of the handheld devices we’re all familiar with, or more use of IoT, more use of autonomous vehicles, more use of drones, et cetera, et cetera. The idea around smart homes, smart cities, smart roads, and so on. You get to this point where you start to recognize that the traditional environment, from a network standpoint, which was designed around shipping things out to the edge and then distributed at the edge via CDN or equivalent, was designed as big down, small up. So for every one meg going down, you had you know some percentage of a meg that could go back up. Well, that paradigm is shifting now, and instead of being the bulk of data going out to consumers, most of the data is now being created with the consumer or near the consumer, whether that consumer is a shop floor or a retail store or a road. It’s being created there, and it’s going to be created in formats and sizes and with latency requirements that preclude its ability to be shipped somewhere else. So, by virtue, IoT gets you to the point of, in many cases, of needing what we’ve been talking about, which is potential for edge computing in one form or another, whether that’s edge compute in a single location for a factory floor, edge compute across a diverse landscape to follow moving work demands of of cars and vehicles, or people, or airplanes, or drones, or whatever, etc. But. That is going to create the need for localized compute infrastructure, which is where edge and potentially edge data centers, or certainly edge infrastructure, if not truly a data center, will be required.
Mark Thiele 15:11
Now that being said, any infrastructure that is close to people or an environment that can benefit from that infrastructure is edge, and so by virtue, a simple assumption about cloud or centralized colocation facilities is that they can still be edge. I mean, I live in Vegas. There’s a giant set of data centers that are five minutes from downtown Las Vegas. Nobody would consider them an edge data center in the traditional conversation, as as you and I would know it, but by virtue of being you know three miles in a straight line from the center of the strip, they are an edge data center. They are a single location edge data center, but they are still an edge data center. So the value is still there. And then beyond that, from a central cloud or public cloud standpoint, there will be some workloads that are enabled via edge information, and those applications will run better in centralized locations with you know centralized management or fewer distributed workload environment. Meaning you don’t have them distributed across 100 different locations, and there will be ongoing demand. I believe at least, for the foreseeable future, for workloads that just have no reason to be at the edge, applications that are more centralized or or extremely compute and storage heavy, and will continue to leverage the cloud infrastructure. So, really, from IoT to central cloud, there’s a potential link for all of them, but in theory, any one of them could be somewhat independent, with the exception that cameras that nobody looks at the data for are sort of useless. But nonetheless, that’s where we stand.
Tim Crawford 16:56
So, what if I’m in a business that doesn’t necessarily have that kind of physical geodiversity, meaning I’m not a Starbucks or a type of organization or business or industry that necessarily has all of these physical points out at quote unquote edge. Is edge still relevant?
Mark Thiele 17:18
I think it’s always going to depend on the business, Tim. And so, if we look at the broader landscape of companies, for every Walmart or McDonald’s or something like that or Starbucks, I’m just guessing. I don’t know that you can find the exact number, but I would argue that for every one of those companies that who is attempting to reach almost every consumer one way or the other, there are another three or four who would also like to hit a major part of those that consumer market one way or the other, but don’t have the kind of presence, or certainly in every town or on every corner that some of these other companies do. That doesn’t necessarily mean that they have any less interest or desire in providing unique services and loyalty opportunities and deeper experiences for their end customers, or even for partners and customers, for that matter, as we’ve talked about recently, even for employees who may now, because of the pandemic, begin working from anywhere. So the ability to create that differentiation and bring more visibility to your brand by being more directly associated with how companies or people interact with your product is a real opportunity for edge. And I would argue that several of the first instantiations of edge for companies that weren’t necessarily on every corner, even though they seem like they are. Companies like Amazon and Apple and Microsoft and Google. You know, you get Google Nest. Nest wasn’t purchased by Google because they needed to be a thermostat company. Nest was purchased by Google. I’m making a broad assumption here. Somebody at Google could call me up and tell me I’m wrong, but Nest was purchased because of its inherent intelligence and the potential ability to create more value between Google and a customer.
Tim Crawford 19:10
Right,
Mark Thiele 19:11
that’s edge. It’s maybe not be an edge data center, but it is in fact IoT. It’s intelligent IoT that ties back to other systems and potentially will tie back to more systems, creating an ecosystem of opportunity for Google to create a deeper, broader relationship with their customers. And whether you’re talking about an an Echo or Apple Talk or any of those things, it’s in the end they work out the same way. So those companies don’t have an Apple Store per se on every corner. Although Apple’s getting better, Microsoft certainly getting better, but they don’t have the kind of visibility that even a Walmart or a McDonald’s do, and they still need to reach their customers.
Tim Crawford 19:50
Yeah, and when you talk about the opportunities that edge brings, I think you brought up a really key point, which I want to exemplify, which is Google didn’t buy Nest. To become a thermostat company, they bought it for the data and the insights that come from those devices, and so I think what you’re trying to say, and and based on past conversations you and I have had about edge, that’s where the value really comes from: is that data and insights, not necessarily the device that’s used in which to get them.
Mark Thiele 20:20
That’s right. That’s right.
Tim Crawford 20:22
So let’s kind of continue down that path and talk about something that’s relatively new, and that’s 5g So you know, 5g is starting to roll out. Where does 5g come into the conversation? What what does it fit in?
Mark Thiele 20:39
I still believe that 5G will be something of a tsunami of change relative to how people interact with technology, and I think we’re only limited in that imagination about its potential impact by our lack of imagination about how fundamental change and how you can interact with IT can change what people develop for you to interact with via IT in a very short nutshell. That’s how I feel about the potential opportunity of 5g But over and above that, some of the potential impacts, you know, positive and negative, are that you know 5g will by itself bring an amazing amount of opportunity in creating new bandwidth in areas that otherwise struggle, creating much lower latency opportunities in areas that struggle, and provide better abilities to separate traffic to provide for guaranteed performance from one customer to another. All of these things are just a few of what really makes sense for for 5g But what is also likely to occur is that 5g will accelerate how data is created and shared at the edge in ways we just do not understand yet. And so, if anyone listening, including yourself, to what I’ve been saying about data creation at the edge and our inability to centralize all that data in any real way, and or get latency-sensitive response to that data from a centralized process. If that were true, with just the explosion of IoT, what’s it likely to look like when everyone’s handheld turns into a fire hose when it used to be the equivalent of a straw, right? From a data capacity standpoint, when stadiums and smart buildings are sending 5g video across long distances, so that they could provide real time same information to customers across broad landscapes. When when 10s of 1000s of cars in a city are running at the same time, using the network and creating terabytes of data every second that they’re all driving. When factory floors begin using video as sensor for almost everything they’re doing, when people learn that they can potentially create business models on top of generic data created by everyone else who’s using data to run their factory floor or determine whether to allow more tourists on the beach, and they create new business models on top of it. It really is, to me, the cusp of a fantastic amount of change. Whether you like it or not is a different question.
Tim Crawford 23:14
Sure,
Mark Thiele 23:14
I think 5g is not a minimum requirement for entering the edge right now, and I think whether 5g is available should not be what dictates you approaching the edge and investigating opportunities for your company, but certainly 5g should be considered part of your discussion going forward.
Tim Crawford 23:32
Yeah. Well, let’s hope that when 5g does get broadly rolled out, it it’s actually a better experience, even from a voice call perspective, you know I can’t tell you how many times I drop a call on 4G or LTE.
Mark Thiele 23:47
Yeah, absolutely.
Tim Crawford 23:48
So, you know, when I think about edge and I think about the data in Insights, and I think about it getting spread across these different components within the spectrum, whether it’s the devices at the edge, or it’s just the simple cloud, corporate data center, edge and edge devices, IoT devices. There, I start worrying about things like cybersecurity. I think about regulatory compliance and privacy. How should I be kind of wrapping my my arms around this? You know, we talked earlier about how yes, this does get a little more complicated, and you got to think about the value of the complication. But what about cybersecurity, regulatory compliance, and privacy?
Mark Thiele 24:35
Yeah, so there are a number of different ways to look at that concern. You know, if you if you think about the device, I would argue that in many cases, the device security threat as an endpoint to your network is one thing. But I think that’s for most devices, that’s likely to be lower on the totem pole for, for risk. As in, somebody’s going to get access to your temperature sensor or your even a video sensor somewhere, and get back into your company’s. Main applications and start stealing data or corrupting data. I think that’s a risk, but it’s a much much lower risk. I think the wider risk is getting your devices to tell you things that they shouldn’t be.
Tim Crawford 25:11
But let me pause you for just a second because there was a situation. I want to say you may know better than I because it did happen. I believe in Las Vegas, too, a casino that was hacked into through a temperature sensor in a fish tank.
Mark Thiele 25:28
Yeah, that did happen. That did happen. And the issue with those things, like the Target store issue, is it takes significant proximity most of the time. But that being said, the reason I position that part of the discussion the way I did is that those are areas of opportunity depending on how you’re deploying. For instance, if you’re deploying on a factory floor or even an office building, you have the potential opportunity to fairly well isolate your set of IoT devices so that they are they could in theory be restricted just to that facility or they’re restricted to access only to a few other locations from that building where you are sending data or retrieving data, etc. And it’s not as if you’re creating open IP addresses to everything on your network. That being said, the onus is going to be on folks to push their vendors and to be consistent themselves with their process and procedures for how these application environments are deployed, right? So I’ll talk about that for just a little bit before I get into kind of another analogy. But when I say how they’re deployed, it’s like anything you deploy. Whether it says it’s secure on the box or not is immaterial. It’s whether or not you change the default password that helps make it secure. It’s whether or not you have a consistent policy and procedure or automation tools that allow you to open and close ports that will allow or disallow unwanted access. That you strip any kind of activity on a network that doesn’t belong there, so that it doesn’t provide confusion or obfuscation for what the network is actually designed for. Sure, relative to IoT, those are all real opportunities, and to help avoid the risk, Tim, you and I have managed large infrastructure environments for much of our careers, and I’m not going to make an assumption for you. I’ll let you decide how to answer, but I can tell you that even in organizations where my employers felt that my team was kicking ass and doing everything that they needed to do to create an environment that was successful, both from a security prevention stand or risk management standpoint, but also from a performance and availability standpoint, there were always times when I just wished for those few extra hours, and what did I wish for those few extra hours for? They would do the things that we couldn’t do because we didn’t have time, and we just hoped no one would notice.
Tim Crawford 27:51
Yeah,
Mark Thiele 27:52
and that’s did we update the BIOS? Did we check whether the firmware on the last disk drives we added in was the latest version, et cetera, et cetera, et cetera? So, what I believe, and just because I believe it doesn’t mean I’ve always done it. I learn mostly from my own mistakes. But what I believe is, if you’re going to make an investment in something, you have to make an investment understanding what it takes to run and own that investment properly. Right. Like getting a pool. Right. Getting a pool or getting a puppy. It’s an investment, and you have to plan. If you don’t plan for how you can auto manage things like security, how you can avoid false alarms, and therefore spend appropriate time on real alarms, how you can maintain to a you know the 90th percentile those things that are maintenance oriented to remove those areas from a risk standpoint. How do those factor into your design of the project and deployment and ongoing ownership? And if you don’t plan for those things, then you’re. It doesn’t matter what you’re deploying. You’re opening yourself up for risk.
Tim Crawford 28:55
And I think that’s a great place to end our conversation because I think that’s a an important distinction to make. It’s not about getting everything buttoned up exactly with every T crossed and every I dotted. It’s about making a value decision. It’s about making a risk judgment as you go through the process.
Mark Thiele 29:17
Absolutely.
Tim Crawford 29:18
Yeah.
Mark Thiele 29:18
Absolutely. Well said. Yeah.
Tim Crawford 29:19
Mark, thanks so much for taking part in today’s episode.
Mark Thiele 29:22
Oh no, I enjoyed it, Tim. Thank you very much for having me. Appreciate it.
Tim Crawford 29:27
For more information on the Cxo in the Know podcast, visit us online@cxointheknow.com You can also find us on Apple Podcasts or wherever you listen to your podcasts. Please subscribe and thank you for listening.
Discover more from AVOA
Subscribe to get the latest posts sent to your email.