Ep 37: News roundup - ransomware attacks, meat, the Russians, Amazon abandons arbitration, supersonic planes, and baby chickens

James Thomason: 0:04

It's The Next Wave Podcast Episode 38. I'm James Thomason here with co-hosts Dean Nelson and Brad Kirby. And our very special guest this week is Mark Thiele, a true cloud industry OG, who I'm sure we give Andy Jassey a run for his money. Mark is currently the CEO and co-founder of a startup called Edgevana with experience spanning over 30 years in various executive positions at major tech, cloud and now edge computing companies, including HP, Gilead, VMware, Brocade, Switch and Edge Gravity by Ericsson. Mark is also the co-founder of an industry organization Data Center Pulse, has made substantial contributions to the data center industry as an educator, chairman of the technical standards committee of the IDCA. In addition to launching his own podcast last year, shortly after ours, he also sits on five different advisory boards in the edge computing space, all while disrupting the cloud industry processes via Edgevana. Mark, it's so good to have you on the

Mark Thiele: 0:51

Oh, man, James, thank you very much, you and show. Dean and Brad for having me on. Thank you.

Dean Nelson: 0:56

Yeah. And Mark, you and I have known each other for a little bit. Like just a little bit. A little known fact for people that Mark and I are related. Yeah, he's my brother in law.

Mark Thiele: 1:06

It's an it's an odd thing that, you know, I'll be talking to me. And he goes, Oh, yeah, you heard what Dean Nelson is doing. And I said, umm, he's my brother in law. And they're like 20 years they've been working with Dean and working with me independently. And they have no idea that we were

Brad Kirby: 1:21

Did you guys wait until this moment to bring that up, because we were talking before the podcast, and you just used to talk about slow play. That's great. I love it.

Mark Thiele: 1:31

That's not to steal the conversation too much. But I can't even count the number of times that Dean and I turned one, and have a dinner table at a restaurant, into a geek fest about whether Sun (Microsystems) or HP hardware was better; what to do with data centers, and just like our families disappeared, unfortunately. But man, it must have been hundreds of times that we did that.

Dean Nelson: 1:55

Yeah, it's really good to know that Sun was better, but I appreciate you confirming that for me, Mark, you know, not that there was any banter or anything back and forth. But yeah, but I think our wives and our daughters, so we both have a daughter, and they grew up literally like daughters, so close. This, you know, single children, but

James Thomason: 2:16

like sisters, you mean, sisters, daughters grew up like daughters? Like they say,

Dean Nelson: 2:20

you know, who scheduled this thing at 7am? Okay, I'm not quite totally on that. But all of them when we get started, you know, they kind of look at us roll our eyes. They're sick of data centers and digital infrastructure and, and all the other debates that we had about how the world is going to be dependent on on technology like this. Hey, but we were right. Right, Mark? You're right. So they go through it. So they just, yeah, we were just educating them. Yeah, yeah. Yeah.

James Thomason: 2:49

A number of other like strange coincidences too, right? You both announced your CEO roles on the same day last year. Did you guys coordinate that? It seems almost impossible to believe. But we didn't. We you randomly selected the same day? Yes. Brother in laws to announce your CEO ships? Yeah. Okay.

Dean Nelson: 3:09

Yeah, it was really funny to me, because I didn't even realize it was the same day, because I saw the announcements, but I didn't look at the day. And until we actually connected mark, and you meant something like, Oh, what? Yeah, wait a second. But that was that was neat. Because you know, the parallels, we talked about this with, you know, we started data center pulse back in 2009. And that was all just really to kind of have a community come together, because we really didn't have one for the data center industry. You know, there wasn't really a way in which you get people together. And it grew to over 9000 members in 66 countries pretty quick. And that was fun. That was a blast for us, to kind of get people together.

Mark Thiele: 3:45

There really was to the point, the lack of a voice, which is the original, I think Dean actually came up with that statement, provide a voice to the end user community of the data center. But there were two common things happening there. One was I mean, and to some degree, they still happen. One was the vendor own the conversation on what to do next in a data center. And the average data center builder had no idea what to ask a consultant, and they just said, Oh, we need a data center, and the consultant would decide for them what they got. And so those two things meant way high cost and inappropriate builds 95 times out of 100.

Brad Kirby: 4:19

Yeah, Accenture made off well, and I'm sure many companies did. Every consultants didn't. Yeah, every consulting. But yeah, I mean, I come from a consulting background. I haven't done too much in data centers. So I'm new to this space. But over the last four years now, I can't think of two better people to be representing the global data center industry than the two of you. I know you're both quite humble and would never say that about yourselves. But it's it's a privilege to have you both on here.

Mark Thiele: 4:49

I mean, except for the fact that I'm better than Dean. We're very humble.

Dean Nelson: 4:52

Yeah, yeah. He worked at that little company called HP. I don't know what Yeah, before they even had an E in their name. Think about it like that. Was I don't know. I know fledgling?

James Thomason: 5:02

Yes. Yeah. It's been weird time. He got a divorce. Dell got married, you know? True. I'm not sure these are strategies so much as just machinations of the sandwich

Brad Kirby: 5:14

Deloitte's building apps or the PGA Tour, like, tour. yeah,

Mark Thiele: 5:18

Yeah, that's a podcast all by itself. James's is how many companies have done org changes or acquisitions as strategy for growth and improvement as opposed to actually just changing how their business operates, what their culture is, which is where it all actually needs to start from. God, I mean, we could make a Harvard business course out of that I think.

James Thomason: 5:40

We probably should, because I think acquisitions and divestitures they get they get billed a strategy when mostly, they're just financial engineering. You know, I mean, at the end of the day, that's, they're less about innovation or changing how companies work. They're more about like, well, we don't like the way the balance sheet looks. You know, like our we don't like our forecast. So what can we do?

Dean Nelson: 6:01

Right, right. Yeah, most of the times, oh, it's the we're losing in the space, and we need to go back and accelerate. So we're gonna go acquire a company to try to boost up or add add something to a strategy, right? Mm hmm.

Mark Thiele: 6:12

Yeah. But then, you know, you think about, like HP's acquisition of Compaq. And in real terms, very little in the way of new innovation, it was just buying more customers in the same market. Well, if your market is already struggling, and all you're doing is buying a bigger piece of that market, it doesn't change the end result.

James Thomason: 6:30

Right? When I die and go to hell, they're gonna put me in a room where I have to install Windows NT on a proliance server over and over again.

Mark Thiele: 6:38

Yeah, yeah. Well, for me, it was in trying to install 40 MB's or 38 MB's on a 40 MB hard drive on a Windows 386 box with Windows 3.1, and trying to figure out what files to delete so I could create swappable memory space.

Brad Kirby: 6:55

My first data center acquisition was actually in 2014. It was at Brookfield, and it was their first data center, tiny and all they did was it was like an Ontario company that hosted specifically just for Microsoft Azure for Microsoft Dynamics, just for the ERP. That's it and I didn't understand the play. So I learned about it from a very isolated level at that point, but now Brookfield's one of the largest, or they're becoming one of the largest data center owners through investment, just as an asset manager. Right. So it's been interesting to really follow the the evolution through.

Mark Thiele: 7:34

No, it's been amazing. I mean, you guys probably all saw the the almost$10 billion that Equinix has raised to pursue building hyperscale data centers, right. Yep. It's just a crazy space right now money wise.

Brad Kirby: 7:48

Equinix announced a partnership partnership with Nvidia yesterday. Oh, yeah. Yeah. It's moving to the edge actually, yes. Yep. Yes, I'm

James Thomason: 7:58

What I'm about to say it's not investment advice. But I'm wondering like, how many of us feel that datacenter reads are a pretty good deal, like in terms of the their potential to earn consistent returns over the next decade? I mean I don't see datacenter growth slowing down anytime soon, do you guys?

Mark Thiele: 8:14

No. I don't think so. I think it's like any company, somebody cannot innovate appropriately in order to stay ahead of the curve. But realistically, we haven't even touched real edge development. yet. We haven't really seen the impact of 10 or 15% of global companies becoming digital equivalents of their former selves. And what that means to new data center requirement and hardware purchases, etc. I have no doubts that data center growth will continue on the pace combined of cloud and colo and private have somewhere between 15 and 20% year over year for the next 10 years.

Dean Nelson: 8:50

The numbers we're seeing, there's a forecast of at least 10,000 megawatts over the next three years alone 10,000 megawatts globally. And you just think about the emerging markets APAC's going to explode, because there just isn't enough capacity there. And you think of EdgeConnex and Adani group, right. Adani is massive in India and the amount of money pouring behind that they're going at India, just with data centers, and then just any other part of Asia, but also LatAm. You know, I've seen this in all these different conversations that, you know, the infrastructure isn't there yet. When if we think about Core Data Centers, they're going to leapfrog I think into a lot more edge deployments, as companies are starting to go forward and say how do we expand and and get into these specific physical Metro markets? Alright, so there's a big differentiator here, I think kind of like India, also where they leapfrog and went right to sell, there's not a lot of landlines and things, they were starting to use wireless, to be able to get people connected in a different way. So a lot of these markets are doing that. But just in the core market, just across the United States. I mean, think about the amount of growth just in Virginia, again, the center of the Internet, and then you think of all the other markets that now are really established- Phoenix - I've said this for years. Now all the cloud players have finally arrived, there are regions that are going into right into all across Arizona. So here you've got a hub, and then you think about what's going on in Texas, and then up in the Chicago area, and all of the United States is just growing these big, big cores. But, you know, Mark, we should talk about your company, and how the big players are a big deal. But what you're actually putting together is now leveling the playing field for a lot of others, as well as the tier two markets. Is that correct? No,

Mark Thiele: 10:33

it's true. I mean, I mean, a supposition, assumption, guiding principle, whatever you want to call it behind the creation of Edgevana was the notion that the best way to get people into new locations was to leverage what was already there. Right now, Dean's done more data center work than I have. And so I'm certain he would agree with the comment that whether I'm looking to build 100 megawatt campus, or a five megawatt building somewhere on the edge of Fremont, California, the due diligence upfront before I bulldoze, the first blade of grass is roughly the same. So if I'm Amazon, or Equinix, or Digital Realty or Alicloud, and I'm thinking about edge solutions, I can't swallow the economics of saying whether I'm building 100 megawatt campus, or a one megawatt facility somewhere, I still have to spend $3 million on the due diligence to get there. And that's, that's just not economically feasible. When you start talking about the kind of distribution that we're talking about. I mean, companies say, oh, we're going to build 200 locations. Well just put that in relationship to what a tower company has somebody like ATC, for instance, they have 200,000 towers in like 80 countries, right. So the best thing we can do, I believe, from an Edgevana perspective, as Dean pointed out, is level the playing field for helping buyers lower the barrier to entry to get access to data centers where and how they want to use them, regardless of whether they're buying from 100 suppliers or one supplier. So when you when you change how you look at the market, instead of saying, I've got these five big companies and I can make the pragmatic choice of going with one of them because they get me to 60% of what I want, or looking at the 5000 other independent data centers and say, How can I turn those 5000 data centers into my network,

James Thomason: 12:25

We should back up for the audience just for a moment and talk a little bit about what is happening in the market. Perhaps first and foremost is to my eyes and ears, the industry has gone from "edge isn't needed", to "we're late building the edge" in about a year. And Mark and I we we both started edge companies kind of early in terms of the market emerging. So what Mark What do you see, I think we should tell the audience a little clearer, like what that's gonna does. And maybe we can start just by saying, what are the key use cases you see emerging an edge? Like, where are they emerging? And what how does edge vahana fit in? How do you help companies? Yeah, I

Mark Thiele: 13:01

The most common, we're seeing a lot, right, we're seeing probably 10 different use case models that are better minimum, but the top two or three include retail, gaming, and manufacturing, closely behind that would be entertainment, logistics, or supply chain oriented. And then probably healthcare, right. And, you know, depends on who you talk to any one of those might go up or down on the list. But generically speaking, that's what we're seeing. And the demands of the folks building edge in those environments tend to follow three categories of need, right, with the assumption of creating differentiation, right? The biggest part of it is they're trying to create differentiation from how others are fulfilling the same need in the market, right. And so differentiation is enabled through things like latency, the ability to react to data instantaneously, where the data is created, and the ability to supply that service, with or without connectivity to the rest of the world. Right. I mean, a simple one, I don't even normally lump this into edge. But think about a mining operation out in the middle of Canada somewhere or oil rig in the middle of the Gulf, they need edge capability. the very reason they need edge capability isn't always because they need a response to machine issues in one millisecond. But it's just as important that they are able to continue to operate regardless of whether the connectivity is big enough or stable enough to enable work from some other location. Right. And so when you think about a retail location or a gaming service, the same thing applies the change from traditional deployment to a campus versus changing that campus to 1000 or 10,000 endpoints is that with the campus, if I'm enabling the campus for the company I work for I've got three or four access points that guarantee resiliency to the network and I can do that, because I'm focused on one area with a lot of people. But when you're a company like somebody like Walmart, or McDonald's or Chick-Fil-A, from a retail standpoint, you can't double your networking and quadruple the capacity at 11,000 or 30,000 stores, that's not even feasible, let alone cost effective. So building edge to be able to operate independent, is a critical aspect of being able to say that not only can I offer differentiation, because I'm providing customers with new experiences, deeper level of experience with my product and services, better integration with with what the customer wants, and what the company is designing. But I can also ensure that I can guarantee 24/7 operations with these new services.

James Thomason: 15:44

And this is a really interesting and a point that gets overlooked a lot when people talk about edge. And like why the edge is, as you said, we often talk about latency and bandwidth and the size of data and all that. But the other sort of macro trend is that things are moving out of the realm of abstract stuff like spreadsheets and web browsers and into the real world where machinery is involved human beings are involved. People can get hurt or die. Or maybe you're trying to prevent that with the application in question. Or maybe simple things like you don't want your vial of vaccine serum to drop below a certain temperature. And even to make that make sure that's true all the time. And someone forgets their clipboard on their rounds. Now you don't know all those kind of cases, necessity, directly proximal computing infrastructure, because you can't rely on the network to be there even at a high level of availability. And you mentioned healthcare, and that's another one where am I talking to healthcare CIOs at some very big campuses in the United States, they have really small SLAs, where, let's say that they take an X-ray or an MRI, they have a 30 second SLA to have that imagery show up on a doctor's terminal, and you're down the hallway. And that can be in like critical care situations and others, but what the these big University Hospitals etc are doing is they're writing now their own imaging code, which processes the raw data off of the imager, right. So if it's an X ray, they're able to better detect conditions like cancer or vascular abnormalities, using algorithms that they've developed in house versus in a previous decades. They've just used whatever the vendor of the machinery gave them. Right, right. That's a real. So those that imagery could be? Well, it is dozens of gigabytes, right. But they still have to do that within that 30 second SLA. And so again, remote connectivity, doing that in the cloud as a non starter, right? It's it's purely not going to happen. And the CIO, of the institution, I'm thinking of back when I was trying to sell him cloud was telling me straight up like, we're never we're never gonna put that into Cloud. It's not gonna happen. Yeah, no, I mean, it was the other way.

Mark Thiele: 17:52

Yeah. Now you bring a really, really important point to add to what I said, when you first asked me the question. And, you know, I talked about the idea of not being able to accept the service being down. Well, that's an inconvenience. Now, it's a money losing inconvenience for average retail store, but it's an inconvenience. But when you take healthcare, or a manufacturing floor, or some complex environment, a chemical facility or something like that, what happens, I think, which aligns perfectly with what you were just describing, is you get to the point where these systems are replacing human consideration from a safety and protection standpoint. And now, effectively, if those systems aren't working, humans can't work in that environment anymore, because they no longer have the capability to provide for their own safety. Because the machine the edge environment was doing that. So I think it all boils back to something I wrote about a couple of months ago on my LinkedIn blog, which is that, you know, the downhill momentum of edge will really begin to accelerate once a number of businesses have become even partially dependent on the services that they created in these remote locations, whether that remote location is in in a mall somewhere or in a smart building or on a roadway or in a hospital, once they become even remotely dependent on those things. The accelerated requirement for enabling and protecting those services is likely to become more visible to the average industry watcher,

Dean Nelson: 19:21

we talked about the pandemic earlier, and just how there's a corollary here with the reality of what edge and just the digital infrastructure is about meaning that people thought that the, the pandemic was kind of a blip. And if you think about all the data that is generated than the amount of remote work that was done, it's not the Blip. It's not a surge. It's the new baseline. And then if you look at the reality of it, so we've gone to these predictions that would be 44 zettabytes of data that was going to be generated in 2020. Remember that thing from IDC going from four zettabytes to 44 a 10x growth over that period of time. And when you look at the reality it was 64 zettabytes last year. Which means that 175 zettabytes is going to be by 2025 is going to be over 225 zettabytes. And I mean, that's a mountain of data. And if you think about where that data is what you just walk through mark, it's generated to certain places doing critical things for business reasons. And the volume of data can no longer be sent back to these places these call centers to do what they need to do. And then the criticality of the work that's going to be done at the edge. Like you said, if a manufacturing plant shuts down, or a utility shuts down, because they don't have the ability to manage those things, during the crisis, or just during whatever kind of business action is significant impacts. So I think the reality that you're talking about is that there are leaders and followers in this space, and the leaders are the ones that are realizing that I've got to be able to do this locally, the followers are going to find this out in a few years and then scramble to get there. But there's huge opportunity for I mean, really, all of our companies, if you think about it, we're trying to go back and put the baseline here for about the future is going to be you're building a distributed edge platform, James, right. And Brad, with EDJX, your distributed platform Mark that's going to be able to allow you have access to all that capacity anywhere you need it whenever you need it for whatever business need you need. And then on VPS, we're building up this infrastructure to make sure we optimize everything because cost is going to be a big deal. availability is going to big, big deal. How do you distribute across all of it? And sorry, I'm pontificating here. But it's interesting to watch how people kind of get there eventually. And so Mark, think about specifically, the question I have for you on this is you and I've struggled with this a lot that you know, it's like we can see it. So clearly. That's what the future looks like. It's like this. And then other ones are the naysayers. I mean, James, how many times have they tell you that like cloud is nothing? Right? Why would we do that we have all these things. So what how do we get people to accelerate their thinking around this topic, versus us beating our head against the wall with a lot of people?

Mark Thiele: 21:52

Well, I think I advise a lot of companies, as was mentioned earlier at the beginning of the show, and one of the common messages and discussion topics that I come across with them is that if they're in a position where they feel like they have to educate the customer, or in many cases, educate the VC, that can be a problematic trail, right? It doesn't mean it's impossible, but it's a problematic trail. And right now, unfortunately, as an industry, we are still educating the populace. And one of the things that I said, I don't know, there's another podcast or a webinar or something that I was on with somebody asked me a similar question to what you just did, Dean, is that, where should people look for edge? And I said, Well, you guys don't realize that edge is already solving for 1000 different things on the market. But people have just done it. It's not hit the airwaves, it's not, you know, nobody, nobody's coming out and said, Oh, I work at Lumen, or I work at Lumentum, a good friend of mine, CIO at Lumentum. And we've done edge, and even 5g on our factory floor. Nobody's coming out and saying that that's just a natural progression. Where in the industry, did the world get flooded with edges here, because chick fil a did to nukes and a complex implementation of open source software in every one of their stores. Right? So edge is already here. It's entering the market in different ways, and not hitting the market in in ways that everyone is looking for. But edge is already here. I think, Dean, to your point, I think the pandemic has accelerated that story significantly, if I had to guess I would say it's accelerated by about two years. I think our best bet not only for our companies, but for the industry as a whole, is helping people now come to understand that the technology isn't the issue anymore. Not even the cost, right? One of the main things that I try to get get across to people as if they think edge is premium, and therefore it's too costly. And I have to have an enormous return on investment before I can even consider deploying to edge, then you're looking at it wrong. Right, you're carrying existing assumptions to the edge, which is not the way to think about it is that we have to help unfortunately, we have to help the world understand that this is a real area of differentiation. And it can apply to almost any part of your business, from the shop floor to accelerating efficiencies within your business to better connectivity or connectivity, technical away with a connection to your customers deeper involvement with your customers. And as Dean pointed out earlier, the companies that use these technologies now are going to differentiate and make a difference and their businesses will thrive as a result also to kind of tie this up and you know, get off my stage is that digital transformation by virtue of what the single most important part of digital transformation is, which is better connection and interaction with the customer in relation to how you do business and how you work with your customers. Edge is a natural opportunity space in that transformation. Edge is effectively the way to change from a flat web interface, phone trees, emails, to the ability to provide instant response to customer interaction with your product or service, to putting the customer into your product management roadmap to creating experiences for your customers that beat the competition. And I'll close on one last point is that a lot of people keep telling me "Mark, but you know, the average customer doesn't care if it's 400 milliseconds or 100 milliseconds." And I said, Yeah, they don't care, until someone figures out how to do it for 100 milliseconds at the same or less price. Now they really frickin care. Yep. And you have a competitive advantage. It doesn't matter whether that competitive advantage is a true real value in the sense that it saves the customer money or something. The fact is, is that I want to use the interface that's more intuitive and faster and entertaining than the one that isn't.

Brad Kirby: 26:08

Yeah, and I guess just from my perspective, I have one example that I think I like to use for with people. And it's around, it's around digital twins, really. And the concept is a bit of a misnomer, right? It's really just a digital replica of an object. It's not a human being unless you're thinking, you know, the movie Dead Ringers starring Jeremy Irons (or West World). Yeah, West World as well. There are examples of that. And for sure, but a digital twin is really just a digital replica replication of an object. And the first example is actually created by the term was created by NASA, right? I think back in the early 2000s, even though the first digital twin was also built by NASA. And that was during Apollo 13, back in the 70s. So if you look at that, and the concept of recreating spaceships, it's really to enable decision making so that you don't have catastrophic errors like we've seen in prior missions. I think that your daily life, let's use Google Maps for as, for example, right, in a way, it's creating a digital twin of your streets. But still, I was stuck in traffic in downtown Toronto for 30 minutes. And there was a road closure, and there's like a for 30 minutes on one stretch of road, there's a crane sitting in the middle of the road, no signs until too late. And meanwhile, in Japan, I'm getting told which car to get on to because I know how many people are in that car at what time of day. And what's the best car to get in to optimize my exit, this is your stop to make sure you catch the next train. And like that's such a basic 2d example, right? Now get into geospatial, 3d, drone delivery. And you're opening up a whole can of worms in terms of the amount of data that's required, the amount of sensors that are required, and you look at history and like electrification, and I'm quoting one of our our other guests, but Jeff DeCoux. But that's a great example that if we don't think about this, now, if we don't plan for the edge today, you're just gonna get this mixed like this, you're gonna have massive, massive security holes everywhere around the cities. And we need to plan for the future today. And then I think digital twins are a big part of that. I think that concept is important. That's one only one aspects. But I'm fascinated by that. At the edge.

Mark Thiele: 28:30

And, Brad, I mean, you bring up I think, an excellent point that maybe even gets missed to the casual observer. And that's that, what is one of the most important areas of revenue for a large city? Tourism. How much does your tourism have to go up? Because people enjoy interacting with your city? In order to make up for any spend? You make? Not very much. A few 1000 extra people, a few $1,000. It's huge.

Brad Kirby: 28:59

I'm thinking the augmented reality, we haven't even seen, we haven't even touched the surface of that yet. Because really, the compute capability at the edge isn't there to create that that replication, right, the cloud is not gonna be able to serve that because of latency.

Dean Nelson: 29:16

Brad, you you bring up anotherpoint in that. So think of Mark, you said that, what does it take for a city to actually attract more people? The simple thing of access to things or just transportation. I mean, not even the cool like meat stuff with engagement, etc. But just like, it's easy to get around the city ready to set, it's like, oh, I've got this thing that makes trains, you know, predictable and easy and not, you know, so there's, there's a huge amount of cases just in that. And I look at the latency aspect we talked about before and inside of the cities themselves. There's a chicken and egg. And you talked about retail mark. And so here you've got all these things, whether it's a coffee shop or a restaurant or something, they've got all these places in a highly concentrated places with people and they can't justify why they would go invest in edge in their store? I don't have a use case that would actually justify why we put compute there. Why would I need to do this? And so they don't do anything because they can't justify it? Well, what if you flip that back around and take this use case like the app store? This is the App Store came out, developers started throwing things on it, and not to tout EDJX, but if you look at all of a sudden you have a platform that allows you to experiment lots of different things, what would you do with that if I could go back and use compute 100 feet away? Or a mile away in the same city at that kind of latency reduction versus going back to the cloud in a few 100 milliseconds? What would I do with it? And and I don't think we will allow that for the developers and the product experience people and all that to say, by the way, here's a new playground, go to town. If you did it, what would you actually do? Mark, you're talking about differentiators. If you did that, all of a sudden, you would have an engaging experience and new delight for your customers and things that you hadn't even thought of before? Because you're bound by the limitations of the way architecture is put together. Right for the last 15 years?

Mark Thiele: 31:06

Absolutely. Yeah. I mean, I think if I can be so bold, I would say that a blog I wrote about a year and a half ago, called the edge marketplace, corresponds with exactly what you're talking about Dean, and that's that, when edge will really explode is when somebody like Dean, I would say Dean, because Dean actually knows how to code. James, you know how to code I, I couldn't code myself out of a wet paper bag. But over a weekend over a week with a couple of friends, when you can develop an app that's meant to help customers navigate a shopping district or see museums in town more effectively. And you can actually deploy that anywhere, at almost no cost, until it gets used, how many customers actually have to use it for you to make money, when you can get access to a billion or 5 billion customers. It's the exact same principle as the Android store and the iPhone store. We didn't go from having 30 apps for our phones to 2 million, because everybody just built apps and gave away their time and money. It was easier, the barrier to entry was lowered. And this allowed people with even the dumbest of apps to deploy them into the environment. And that's when I really see the edge market exploding, is when people can do exactly what Dean was describing. They can look at an environment and whether they're actually the coffee shop employee, or they're just an IT observer of the environment, they can contribute to the experience in that space.

Brad Kirby: 32:35

Yeah, I know. That's it. That's definitely James's vision. When we first started talking three years ago, when when EDJX was founded, or at that time, it's just the time to development, there's so much when we move to the edge of the DevOps orchestration, and all the work that goes into that. It's just it's just a hindrance to development. And it's just going to delay the process of rollouts and production issuances as you bring it into the real world. So there are challenges and or that's what we're all trying to solve for.

James Thomason: 33:06

Yeah, it's it's interesting to me that these patterns in technological cycles and trends like this emerge over and over again, it's almost as if, at the emergence of every new, major technological shift to the detractors of technology stand up and say, isn't the world good enough? The way that it is like we already have horses and buggies. And, you know, we can get around town just fine. Why do we need an automobile? And by the way, automobiles run on gas and you can't buy there's like gasoline infrastructure across the country? How are you going to get to California, from the east coast, I mean, you can't do that, and the horse eats hay, and that's natural and better for the environment. And we shouldn't have automobiles, because they're a bad idea. And I feel like we encounter this every single time, there's a huge shift coming in the way things are. And it's partially because it's so disruptive. But it's also because it's hard to see how things need to change when there's no possibility of them being more efficient than they are with current technology. And that's trying to put a fine point on it. As technology evolves, when it's successful in a market economy, it becomes more and more componentized. That is to say that suppliers enter the ecosystem to supply the widgets, if you will, you know, that are used to produce said technology, but also to produce the things that run on said technology so that the cloud is an example of that, but maybe a little more practical. When you go to the store, Best Buy or something and buy a Gizmo you know, give any thought whatsoever to like higher in a power or connect with that thing. You just take it home and plug it in. And that that is the benefit of componentization. That's the benefit of platforms, right that you begin to interoperate in more or less a standard way and you take lead to take for granted. And when you take for granted, it leads to the emergence of a lot of things and I mean that you know a lot Products a lot of different services new things that they get created. And so people ask me well, where are you know James? Where's all the demand for edge gonna come from? We have the cloud already. We have whips and buggies. I mean, we're fine. The cloud scales the cloud works in the right amount that it does. It's it's actually what it does. But marks earlier point, the edge is already here, you just don't see as because the way that people do things, they're coping in different ways, right. So there are tons of manual human what I call like clipboard processes, people running around with clipboards, doing things that are not a good use of human beings time, or not a good use of a corporation's money, and so forth. And when you start to really peel the onion on it, you begin to realize that like, all of these industries, haven't modernized at all, like zero. Even in medicine, like the example I was pointing to earlier, a person with a clipboard comes by and checks the refrigeration to make sure that the vials of the vaccine have not risen above a certain temperature. They do that on their rounds, you know, and then there's no immediate way to know whether that's happened or not. It's all it's all human process, or in a food and beverage setting, like in a retail or food and beverage setting. Like what what is my current inventory in the store? I get a clipboard in person running around, or I'm selling food in my restaurant, what are the temperature, my coolers? What's the temperature of my fryer? How much stock do I have? What do I need to buy? All these things are more or less human processes. And so right now, I'm sitting in Miami, and I'm here at the IoT evolution show. And IoT was so dramatically hyped, maybe five years ago, I mean, you couldn't, you know, people were breathlessly breathing, the IoT and, you know, at every opportunity, and it really did get overhyped, that we hadn't even scratched the surface of transformation yet created by IoT. And that's when we started EDJX, three years ago, we realized, that tell people all the time, I told my wife like, I will absolutely never start another infrastructure company. That was enough. I'm going to do marketing, never, it's way easier. infrastructure is hard. And you know, then we realized, okay, the biggest, I think all through all of us on the call here have been fortunate enough to see several different versions of the internet starting at the Al Gore, Internet, and then the.com, internet and the mobile internet and cloud, and now truly the the Internet of Things. And we don't even know what's possible, because you can't like at the far edge of the network, you can't take anything for granted, there is no connectivity, there are no computers. So it's not even obvious in a way, like what what is possible, until you put the platform infrastructure in place. Right. So there, there are so many more. And it really does feel to me, like a similar moment in time when, when Apple released the iPhone, you guys were talking about the App Store, okay, and they were putting a computer in everyone's hand for the first time, literally in their pocket or in their hand. And no one was even sure. What, what this would be used for exactly. I mean, you can play music, it was your camera, you could check email with it. But that that's to say nothing of how many apps have been written for the phone. And that's because the phone is a platform and I. So I see the same thing happening now and just starting to happen really in the Internet of Things. And that is the need, right, the need is to allow the platform effect to materialize, if you will. And then the proliferation of use cases happens on top of that, if you were on it, not to say if you build it, they will come but if you build it, they will come because they're already there. It's not like you're inventing the edge for the first time the use case is sitting there, you just, you just don't realize it because you pay 500 employees to run around with clipboards all the time. Or like crazy stuff like in the oil and gas space. When they want to know how much fluid is left in the tankage. They send a guy up the ladder in the scorching desert, you know, West Texas heat to open the hatch on a tank of caustic chemicals and look down in there throw a stick in a rope and like Well, it looks looks like it's about a third full to me. I mean, that's how they do it when they look down to the markings inside of the tank. That's crazy, right and dangerous. We shouldn't be sending people. We send people in pairs around the chemical plant. If one of them drops dead. The other one runs back and says we got a leak. You know? Yeah, we're Canarian people.

Mark Thiele: 39:22

This is great. It is crazy. And you know, three things associated with what you were just describing James. The first one is about the market as a whole, right? Let's say the four of us are it's 1900 1920, give or take a couple of years and we're on a boat fishing in the Gulf of Mexico. And Brad just happens to be a geologist and he happens to know through some intelligence, he owns that there are 10 billion barrels of oil underneath the bottom of the bay. And we all sit there going it's too bad. Nobody will ever get it because there's no technology to drill underwater. Right. And we just walk away gone. Oh yeah. 20 billion barrels. Whatever it is, we're just gonna forget about it. That's what the market is today. And so many of the people that I talked to, are those people on the boat saying, Oh, well, it's gonna be hard. So nobody will do it. No, this is a trillion dollar market opportunity. It doesn't matter how frickin hard it is. People will figure out the pardon my French fuck out, because there's too much value there. Right. Now, the the other part you were talking about? And I think we've all related this in one way or the other. But edges, like the solution for the world's equivalent of a scene in Black Hawk Down. Are you guys familiar with that movie? If you haven't seen it? Really, I think it's the best war movies ever made. Right now. I agree. But there's, there's a scene and I read the book before I watched the movie. But there's a scene in the movie where they're trying to get back to base. And the communication between guys on the ground, guys in the air guys back at base and then back to the air and then back to the ground meant that they always pass their turn before they could execute. Imagine that when you're talking about like, another comparison is this drives for the average human thinking about how fast the IO was between CPU and disk? Like, how could that possibly matter if it's one nanosecond versus two? Well, it matters when you go from five transactions to 5 billion transactions, right. And so edge from my perspective, whether it's in a hospital, whether it's in a city, whether it's on a manufacturing plant, once you start recognizing that I can take out 10 nanoseconds there, I can take out 20 milliseconds there. And I'm doing those things on 10s of 1000s of transactions an hour, all of the sudden, the idea of actually steering with accurate real time information becomes a reality for the enterprise. Right? And we've been, I mean, anybody here that's ever talked to anyone in business leadership, from an IT perspective has always heard we need real time data. Every single time I ask an audience about that they own everybody that's been in it says, Yes, we've been asked for real time data. And then I follow up and I say, how many of you believe the enterprise has actually ever responded in real time to real time data? All the same? people raise their hand again? Nope, they haven't. Right? And this is that opportunity for my perspective. Agreed.

Brad Kirby: 42:27

And I think it's the time is not only just the milliseconds and nanoseconds, it's the it's the amount of efficiencies gained from fast decision making. If autonomous decision making right like he like with machine learning, and whatever you want to call it, machine learning, automation, digital transformation, AI, it's all in the direction of giving us more time to think and create and innovate and compete. Now the thing is competing with our another country, neighbors, maybe not neighbors, but maybe on the other side of the world. China for one, right? I'll say, with China there's competition there. And if not even an informational war that's going on.

Dean Nelson: 43:06

They're definitely doubling down Brad. You're right, yeah. But it's, there's progression happen all over. And now, Mark, let me wrap this back around. Because we keep talking about the, we can clearly see what the future looks like all of us are investing in companies and opportunities to take advantage of that future. And it will come, right, but it's going to be this, this forcing function that's going to happen. And I think it's going to, you know, we talk about cost, there's going to be a cost reduction, but there's also going to be this acceleration of experiences when it comes down to the performance at the edge. And that's going to lead to business opportunity. Everything here is how do we take advantage of that business opportunity? So I want to go back to an article that I saw that he wrote an MIT article with a few other folks, because it starts to highlight the arguments against right for edge computing aspects. He just give us a quick touch on that. Because these are the things that if people aren't naysayers to edge, and they're listening to this podcast, these are the elements that you should be considering about what could disrupt you if you don't really think about it for what the future will be.

Mark Thiele: 44:09

Right. the article is a great example. It's actually an article written by a gentleman by the name of Blessin Bergasi, and it's really kind of an academic article around edge. And I was a literally a bit contributor to this, the vast majority of this article, and the research that went into it was done by Blessin, or maybe some of the other people that contributed to the article. It's interesting to note that his first effort, had writing an article was actually proving that public cloud would solve for all this because there was enough distribution. And by the time that was his real hypothesis, right? By the time he was done with his research, he came out with an article that said, there's no way to solve for the potential opportunities and needs of what is considered to be the edge Today without edge infrastructure, and it's ironic, I mean, I'm sure all of us have seen it. But and if you want to see this article, you know, go to my LinkedIn, you can find it there. But if you think about what he's saying, and if you read the article, you'll find this, he's actually done the studies to make this a deterministic outcome. First and foremost, he's independent. He's not me, he's not Dean. He's not James, he's not Brad trying to sell a company. I mean, he's completely independent, but resemble that. He comes out basically saying that all of our suppositions historical suppositions about either one, latency doesn't matter or to latency to the cloud is good enough or wrong. If I can paraphrase the whole article. That's effectively what it says. So I really urge folks to go take a look. And I don't know if anybody else on the podcast today will post it. But if you if you want to go to my LinkedIn profile, you'll find it there. As a posting from earlier this morning, it really does put in perspective, the difference between you know, it's like James and Dean and Brad and I were talking about edge computing, and we're talking about 1000s, we're talking about 10s of 1000s of locations. And people think that Well, Equinix is big, DRT is big, Amazon, Microsoft, they're all big. But if you added those companies that I just mentioned, and two or three, the other biggest players with data centers in the market together, they come out with about 1500 data centers, you know, any edge locations, China believes it needs. And this is from 2017 to 2020. Not even to 2021. Just in 2020. They went from thinking they needed a million in 2017. To thinking they need 5 million now. Well, yeah. So the locations the way we've historically been thinking about it, where we make these pragmatic decision Oh, well, you know, this data center company is in is the gives me the best distribution I can find from one company doesn't even come close to offering the kind of opportunities that we've been talking about today.

Dean Nelson: 47:00

Yep. So if there's a stat out there, that there's 7 million data centers globally, data centers loosely made you think about it. And so if we just look at the amount of devices we always talk about, in the past, it was 4 billion addresses that can be done. So then you've got that plus people is about 13 billion things that were on the internet. And James, we talked about this for years, how many things are going to be better than the next? Yeah, thanks to Nat. So now we're talking about Okay, this is going to go from 13 billion to 100 billion. Okay, that was a previous predictions. But now you think about the behind that 100 billion is a trillion sensors. And Mark what you just mentioned, the whole thing is, what is an edge location, what is a data center, all of these things are going to add up into a significant number all over the world. And that data tsunami, there's huge market opportunity in this and all the people that are naysayers looking at that one are gonna get kind of hit by the data tsunami and then figure it out.

Brad Kirby: 47:57

I have one more really interesting article that I stumbled across this morning on AI spectrum. And it was about a backscatter radio that was printed off on a 3d printer by some guys in the UK. And it used literally 1000 times less power, like it doesn't even need a battery. And it can operate at the millimeter wave spectrum at two gigabits per second. And it's a backscatter radio, so it's not an actual like device, but what you could effectively do is put a transmitter on that, and use a 4g phone and connect it to a 5g network. And just just like that kind of innovation is is mind blowing to me, because it's so simple. It's just a very simple device that they printed out with a 3d printer at low cost, saving a massive amount of energy and in effect, creating, you know, solving for some of the problems of 5g. That's just one example of many. And I'm sure we could go on for hours. But that's the type of innovation I think we'll just continue to see.

Mark Thiele: 48:56

I mean, I'd like to leverage Dean's point about the trillion, because that is the common number now that there'll be a trillion devices by like 2030, or something like that. I don't remember the exact year, but it's not like we're talking about 20 years from now or something, it's within the decade. And even if each one of those trillion devices is only identified as providing a few bits per second, add that up, right. And when you consider the fact that video is becoming more and more of a common sensor endpoint, and video is going to 4k and 8k, it doesn't take long to change that average from a few bits to a few megabits per second as an average across that environment. So just extrapolate what that means. The other thing that I think would be a fantastic discussion topic, guys, if you guys want to do this, the next time maybe is there will be a first layer, incomplete, but there'll be a first layer of distribution of opportunity associated with people deploying IoT for the factory for a smart building for a city for farming, for shipping for transportation, whatever, but the potential next wave is multifaceted but largely in my mind, largely around the reuse, effectively the multi tenancy of what's been created first. And so it's not as if Brad and James and Dean and I will be able to log in to somebody else's IoT devices. But we'll be able to share will want to share and create a effectively a trading platform for how data is reused. Because if I'm in retail, and I benefit from the surf aspects of a particular town, that town is already collecting, what the weather is, how much parking is available, what roads are broken, what parks are available, how cold the water is, what the tide is, how many parking spaces are available at the beach, the city's already collecting that none of that data is proprietary. And do we need intermediaries to get that data? Or can we just do right? machine to machine? Or peer to peer? Yeah, but the bottom line is that alone could be a market bigger than the original assumption. Just to getting that first layer of edge out there. Yeah,

James Thomason: 51:02

yeah. Wait, wait till Mark Zuckerberg gets a hold of this. Or someone just rumbled? Did you know you're three blocks from a McDonald's? turtle? That's right. That's right. Perfect example, right, in your reality display? A ton of great points. Your ex girlfriend from 10 years ago is there she goes, she goes online. Know exactly right now put on your sunglasses and run? Yeah, well, we're at the end of our hour. And I just want to mention a couple of things, we're gonna do kind of a different show coming up this week. As I mentioned, I'm live at IoT evolution here in Miami. And I'm running around the show, you know, like I was a reporter and actually talking to people and learning about their companies and what they're doing. And it's, as we said earlier, in the show, it's just very reminiscent of the kind of explosion of innovation we saw around the birth of the mobile platform is happening now and IoT. So we're going to take some of the best segments of that and put them together into a new kind of show, which we'll release later this week, and possibly early next week. And then we're also going to go on to daily show, I believe, the next week or so, and follow up on some of that. So a couple of new new formats coming to the audience do stay tuned for that, I just want to thank Mark for taking the time to be with us here today and sharing his insights. And you know, Mark, I feel like we didn't quite talk enough about edge wanna, and when you guys are doing, so maybe we can just kind of close the show and say, You know how, how's edge vana, helping businesses keep getting wrapped around the axle with this? And with all this edge and IoT stuff? And how are you helping them, build edges and scale and grow to meet the next wave, as we say on the show?

Mark Thiele: 52:37

Yeah, I mean, everybody on the podcast today provides something that is, in my mind necessary to the development of edge, I mean, Dean's company, allows people to leverage the capacity they have in any one location more effectively. Anybody that knows anything about edge, it makes assumptions already about the prime nature of real estate associated with being at the edge. Right. So that's a key opportunity, what EDJX is doing relative to lowering the barrier to entry and providing locations for unique utilization of compute resource and networking around the world. If I could pick just one area related to that, that I am focused on from Edgevana standpoint, it's taking those kinds of theories and applying them to how people buy global infrastructure. And what I am seeing from most of the CIOs that I talked to, and hearing from most of the CIOs I talked to, is that they're tired of creating 17 projects, every time they need a little bit of compute or data center space somewhere else in the world. And I just haven't liked the number 17. It could be 10. It could be 50. But every time somebody that's done my job, or Dean's job in the past, gets asked, okay, we're putting this new application in it needs these new locations, or we think it'll need locations, what do you have to do? There are at least a dozen if not two dozen projects that kicked off for finding new networking for finding the location for determining the values associated with that application SLA and how that drives into decisions about what infrastructure to buy and build, whether it's the data center, the networking, the hardware itself, how's the hardware stack? How much density do you do? How much how important is power versus taxation? The list goes on and on. And most companies aren't good at doing this. So if we're all agreeing that cloud is good, why is cloud good? Well, because 20 years ago, when Dean and I were arguing about who was better HP or sun, when HP one, the common assumption by us was that if we needed a server, we built it. We built it. We frickin added everything it needed to make it work. And then six years later, it became commonplace to be able to just buy an instance of a server in the cloud, right? And yet, if Dean and I had been talking about that, well, we sort of talked about that stuff with some of the things that that son was doing that were you know, Maybe better than what he was doing? Oh, that's cute, etc. What? Okay, I know, I'm happy. I didn't say that. I've erased that from this refrain from that script. But for the most part, most people would have said no freakin way, right? That's not going to happen. It's going to be it's going to be manual, it's better when you do it yourself, etc, etc. So why, if that became true, why shouldn't the acquisition of global resources for your infrastructure requirements be the same? How come they can't be more cloud like, abstract away, you know, data center names, data center locations, who the network is being provided by where the hardware comes from, extract all that away, and give the customer solution. And edge vahana really wants to be able to get to that point with their customers.

Brad Kirby: 55:41

Cool. All right. Well, I THINK DEAN'S got a run. Do you want to pose for a quick picture? And we'll wrap it up? At the end

Dean Nelson: 55:50

Yeah. Well, Mark, hey, thanks a lot for coming in. I think that the democratization of what you're doing, and there's going to be really cool. So we're going to be watching the Edgevana as it grows.

James Thomason: 56:03

Well, thank y u very much, so much. And, yo know, just want to say to the udience that this podcast, of course, is sponsored by nfrastructure masons whose unit ng builders of the digital age, and you can participate by going on the web, to ai masons.org. You don't even uilders of the digital age m ans we're helping everyone C nnect, Grow and Give back, really enhancing educat on opportunities, champion iversity and inclusion a d promoting innovation an technical excellence, and inspi ing sustainability. And if y u don't know much about AI mason, you really should ake a minute, if you're an inf astructure practitioner to go on the web, and check tha out. And of course, we'r sponsored by EDJX, my own com any. We're about two years old now. And we're helping b sinesses by creating a new e ge computing platform that makes it easy to write and dev lop IoT applications using serverless. Computing. And accel rating delivery of content and ata at the far edge of the netwo k. You can visit us on the web a edjx.io. And Mark Thiele, tha k you so much for being on the show. Be sure to check out Ed evana that's Edgevanaa.com rig t, Mark? Yes. Okay, that's ne.com everyone. Thanks so much. We'll see you soon. Thank