This episode features an interview between Matt Trifiro and Ivo Ivanov, Chairman and CEO of DE-CIX GmbH where he is responsible for the global business activities of the Internet Exchange Operator. Ivo has more than 15 years of experience in the regulatory, legal and commercial Internet environment. In this episode, Ivo talks about interconnection, the importance of integrating edge computing, and what technology will look like five years down the road.
This episode features an interview between Matt Trifiro and Ivo Ivanov, Chairman and CEO of DE-CIX GmbH where he is responsible for the global business activities of the Internet Exchange Operator. Ivo has more than 15 years of experience in the regulatory, legal and commercial Internet environment. Joining the organization in January 2007, in recent years Ivo has been deeply involved in the establishment of DE-CIX sites in North America, Southern Europe, the Middle East, the Indian subcontinent, and Southeast Asia, as well as several consultancy projects in Africa, Asia and Europe. Ranked as one of the top 100 most influential professionals of the Telecom industry, Ivo is regularly invited to share his vision and thought leadership in various industry-leading conferences around the globe.
In this episode, Ivo talks about interconnection, how it differs throughout the world, and the importance of expansion especially in developing countries. He discusses how the edge interconnection platform is a missing piece to help enable many advancements and innovations in everyday life, in areas like automobiles and technology. Ivo also explains what he thinks the technology industry will look like five years down the road.
---------
“Interconnection really needs to get deeper into the edge for one very simple reason, quality security performance in general, or a magic world of latency. Latency is extremely important nowadays. I love to say latency is our new currency.”
“Why democratize? Because we have been doing this business for 27 years now. 24 years of those 27 were entirely wholesale industry focused. Just a couple of years back, we started direct business relationships with enterprises, with car manufacturers, with banks, with retailers, with logistic companies. And, these guys, their business is not telecommunications wholesale. Their business is the business of their industry segment. They just want to use this type of infrastructure for security and performance reasons. But they don't need to be experts in wholesale telecommunication. So, we need to make it as easy as possible for them to enjoy the interconnection services. This is what I call the democratization of infrastructure, that every single type of business on this planet can use.”
“I think in the future we need to develop these further pieces of infrastructure and it's not the DE-CIX platform only. We need to see more edge data centers. We need to see more fiber in the rural areas. Middle and last mile investments are extremely crucial also for the economy of the country and serving the people, disregarding where they live and work with the best possible quality of access to digital services. Because if you look at it from a more altruistic point of view, good access to digital services today also means a good quality of life because with economic growth at the end. I'm 1000% sure about this. Diversification of infrastructure is extremely important also for redundancy, resilience, and security reasons.”
“It will be a world which will use more and more digital services as we can even imagine today, further down the path we're talking about the metaverse. We're talking about an internet of a new generation or digital services of a new generation where people will enjoy different virtualization. eCommerce will change, will become extremely augmented and virtual reality driven. We will see in the business world further progress on processing on different solutions for any type of business that's relying on cloud compute, but also relying on digitalization and digital infrastructure.”
“I see a world full of innovation, I see a world full of digital infrastructure and demand for digital infrastructure. Therefore, I do worry a bit that we will not be fast enough to develop this digital infrastructure as much as we can and everywhere, literally everywhere, around the globe. At least, this is our mandate and we will not stop working hard every day to do this.”
---------
(02:00) Getting started in tech
(05:52) Understanding internet connection
(12:19) Edge tech in internet connectivity
(16:47) How interconnection is changing
(21:11) Why interconnection must be democratized
(29:37) Advantages of having data centers in metro regions
(35:41) Interconnectivity in auto and tech sectors
(45:14) The importance of quality assurance
(46:38) Predictions for five years in the future
--------
Over the Edge is brought to you by Dell Technologies - unlock the potential of your infrastructure with edge solutions. From hardware and software to data and operations, across your entire multi-cloud environment, we’re here to help you simplify your edge so you can generate more value. Learn more by visiting DellTechnologies.com/SimplifyYourEdge for more information or click on the link in the show notes.
--------
[00:00:00] Narrator 1: Hello and welcome to Over the Edge.
Today’s episode features an interview between Matt Trifiro and Ivo Ivanov, CEO at DE-CIX International.
Ivo is responsible for the global business activities of the Internet Exchange Operator. He has more than 15 years of experience in the regulatory, legal, and commercial internet environment. Ranked as one of the top 100 most influential professionals of the Telecom industry, Ivo is regularly invited to share his vision and thought leadership in various industry-leading conferences around the globe.
In this episode, Ivo talks about interconnection, how it differs throughout the world, and the importance of expansion especially in developing countries. He discusses how the edge interconnection platform is a missing piece to help enable many advancements and innovations in everyday life, in areas like automobiles and technology. Ivo also explains what he thinks the technology industry will look like five years down the road.
But before we get into it, here’s a brief word from our sponsors…
[00:01:06] Matt Trifiro: I often get asked what shows the best to attend, and one of my favorite live events to recommend is Edge Computing World. It's a great place to network and catch up on the latest in edge computing. They have topnotch keynotes and panels and an expansive show floor. I plan to be there both days this year, October 10th and 11th and Santa Clara.
In fact, we'll be recording at episode of Over the Edge, Live on stage with one of the keynote present. You can get tickets edge computing world.com and as an over the Edge listener, you can snag a 25% discount just by using the code. OTE 25 Checkout
[00:01:38] Narrator 2: Over the Edge is brought to you by Dell Technologies to unlock the potential of your infrastructure with edge solutions. From hardware and software to data and operations, across your entire multi-cloud environment, we’re here to help you simplify your edge so you can generate more value. Learn more by visiting Dell.com for more information or click on the link in the show notes.
[00:01:59] Narrator 1: And now, please enjoy this interview between Matt Trifiro and Ivo Ivanov, CEO at DE-CIX International.
[00:02:07] Matt Trifiro: One of the questions I'd love to ask my guest first is, how did you even get into
[00:02:10] Ivo Ivanov: technology? I got into technology very early, media technology actually in the nineties during my high school times where I started a radio show. The weekend. Really? Yeah. I was at school and started a radio show afterwards, a TV show for near, and this brought me a kind of, uh, into technology.
Of course, not telecommunications, right? Obviously not, but media in generally, This was the first. Kind of a experience done with also digitalization. Digitalization of audio, digitalization of video as the first features came out, et cetera. And usage of content. Of course not in 1995 in the way we do see today, but at least first steps.
They have been done at that point of time. And later on during my low studies, because this is the professional background I have, I focused on specialized on telecoms and media. And where did
[00:03:10] Matt Trifiro: you develop that specialty?
[00:03:12] Ivo Ivanov: In two universities I started in, so Bulgaria, first year of my law status, and afterwards I continued in bond in [00:03:20] Germany.
[00:03:20] Matt Trifiro: You've been with DKS for almost 15 years, right? A while
[00:03:23] Ivo Ivanov: and actually I started my career Dicks as a general corporate council attorney. Yeah. Yeah, that was okay. That was the beginning. I got so excited about the industry itself, about this magic related to telecommunications and the magic around the internet itself.
Then I started learning a lot, the business, the industry, infrastructure, interconnection, data, descenders, bgp. All this internet exchanges got more and more experienced also about, Yeah, the bits and divides of everything around this and over the time obviously changed my career from attorney and corporate council career into a management one and.
The one reason, and this is something probably very special, how I got more into the business side of the industry was the fact that at that point of time in Frankfurt, around 2006 seven, we started approaching Eastern European network operators to connect in Frank. Meeting or interconnecting how we call it today with network operators from the Western hemisphere, US based guys, or guys from Western Europe.
So Frank became suddenly the gateway to Eastern Europe, or in other words, The heart telecommunications heart of Europe where West started meeting East. Right. And I speak Russian on Thep. Bulgarian because these are my roots. So a couple of east new languages, and this was a kind of a [00:05:00] fortunate situation.
So you are
[00:05:01] Matt Trifiro: the interconnector. You're the interconnector of the interconnection. Yeah.
[00:05:05] Ivo Ivanov: I needed to help out with the first meetings in different languages, and that brought me so close to the business. I was excited by and fascinated
[00:05:13] Matt Trifiro: by. You're wearing your enthusiasm on your face. It's really fun to see that.
You know, I come from a software world and so infrastructure's relatively new to me too, and I had that same experience. It's like, Wow. Cause I've always thought the internet is like the cloud. It's this amorphous thing. It just kind of out there, it's bits and bites. Its all digital. But now the internet has physical presence, like it exists in certain places and not in other places.
And data needs to travel across. You know, it's like, it's a whole different way of thinking of the world and viewing. It's like those, they've seen those cross sections of New York City where they show you all the layers and how the system works, but you're just like a, a top driving in your car. Like, it's like, what?
How does this even happen? So, you know, interconnections. Really common term that we all talk about, but I don't even today feel like I completely understand it. So could you help us understand? What interconnection is and why it's important.
[00:06:02] Ivo Ivanov: Interconnection states for bringing networks together. The A side talks to the B side, and we have in general, two types of networks in very general sense, right?
Mm-hmm. , we have the so-called outbound networks, traffic outbound networks. Those networks who do create mostly the content, right? Netflix, Amazon. , Google, Facebook, you name them. All of those networks, gaming networks, all of those networks, they are on the application side, right? In general. And we have the inbound networks.
Those are the networks who do deliver internet access to our homes and offices. So [00:06:40] interconnection is the connectivity between these networks or the way they meet. They interconnect or they get connect. The best way to do it is the direct one, obviously right without Intermediates, where Netflix Network connects directly with a network on the last mile.
Comcast, at and t, Verizon, all the guys who deliver internet access to our homes, to our living rooms, to our offices, but also the mobile operators, right? The guys who deliver 3g, 4g, 5G connectivity, ideally to our mobile devices, and this is what we call interconnection, actually connectivity between networks.
And asking me about interconnection, I must say interconnection is so close related to the term of an internet exchange because the internet exchange or the abbreviation ia, what the company I'm working for stays for Dex now for 27 years and internet who delivers access to the internet to these applications.
They do interconnect. Without intermediates. The best way to describe an interconnection platform actually is the analogy to an airport. As the airport operator is not an airline operator, it's just the operator of infrastructure where the different airlines, they can land their planes and exchange passengers and luggage, right?
So that's what we do for the internet. We. Not an airline. So we are not a network operator. In my analogy, the airlines with internet [00:08:20] networks make sense. So we operate airports with internet where the network operators are the airlines, they are connected directly to the platform and exchange a lot of data packages.
Tons of date.
[00:08:33] Matt Trifiro: So at some point, you know, I, I think about the evolution of the internet and, you know, it started with, you know, 2, 2, 2 computers. One in Palo Alto, and I remember the other one was, it's Los Angeles or something. And the internet connection. Back in the original days, and I'm sort of hypothesizing, but I believe this is true, based on my reading of the history, was the government connecting the networks together, right?
Mm-hmm. . And at some point that was turned over to more private entities, is that when the business opportunity for companies like D Kicks emerge, Was when it became, not so much a governmental regulated activity, but something where networks had to figure out how to connect in a service for doing that emerged.
Is that how that happened?
[00:09:10] Ivo Ivanov: Yeah, quite close to the reality because in 1995, Dicks has founded in urt, established in URT in Germany, one of the, or Dick first platform of Dicks, and has been established with three customers at the beginning. Those free network operators were connected as the first free one to the platform for vocalizing their traveling exchange.
But y 90, 95 almost, uh, close to what you just explained because 1995 was the time where, The commercial internet started to grow, right? But this was the time where the commercialization of the internet creating a more, as a product for consumers, for businesses, et cetera, began to happen. And these were the days where more and more internet service providers, they were established or [00:10:00] appeared on the landscape and with the member of different service providers with products on the market.
Internet access for first emails and email accounts, and very simple browsing the network operators, they realize that localization and interconnection in general, so the desire for better localization, for better connectivity, control, and quality, but also costs is so important. So at this point in time, everything was interconnected in the US in 1990.
So what the guys did is they created a platform in Frankwood where direct local interconnection can happen without sending the bits and bites across the Atlantic.
[00:10:45] Matt Trifiro: Oh, that's interesting. So are you saying, you saying that prior to D Kicks, if I was like the frank equivalent of Netflix and I created all this amazing content that if I wanted to get it to a subscriber in Berlin, I'd have to send it to the United States.
This is absolutely correct. Okay. So when you say local traffic exchange, you mean literally so that Frankfurt can go directly to Berlin or close, close enough of a direct path? No, it's more
[00:11:08] Ivo Ivanov: than we should talk about local, especially if it comes to this organization more in the sense of a couple of hundred kilometers to a thousand.
This is where localization still makes sense. I'm talking about normal applications, but we'll get into this probably talking about the trends and the innovation in the future in the market, but localization in a broader sense. Yes, Frankfurt does certainly the trick for the normal type of applications where we talk about streaming, we're talking about general data [00:11:40] availability for sure, for Germany and beyond Germany.
So in Frankfurt, we do see more than 1,500 networks connected to the platform in Frankfurt in the range of, uh, approximately 2000 kilometers. So this still do the trick for, let's say commodity applic. What do you mean by 2000 kilometers? That means that networks, for instance, access networks from Eastern Europe, they are connected to Frankfurt access interconnecting with applications
[00:12:07] Matt Trifiro: in Frankfurt.
So there are places on the planet that are more interconnected and they're placed on the planet that are less interconnected. And so Eastern Europe is probably less interconnected to the United States. Probably a fair statement. And so what you're saying is that autoimmune Eastern Europe, even though Frank.
May not seem close from the speed of light perspective, it's closer than whenever else I'm going to get my internet. Yeah,
[00:12:25] Ivo Ivanov: absolutely correct. But I just gave the example of the nineties and the millennium mirrors. Uh, a lot of things changed over time as we have now, 2022. We have seen fortunately, a lot of interconnection infrastructure, blacks ones have been created in Eastern Europe as well.
As we know that companies like Netflix, Apple, Facebook, Microsoft, Amazon, just to throw a couple of names on the table, content heavy networks, they also have extended their network presence, their point of presence in Eastern Europe and beyond, right? Today. Decades, operates almost 40 different platforms on four continents.
In the US we have five different markets. New York, Chicago, Dallas, Phoenix, Richmond, and more to come in Europe. We operate 14 in the Middle East. We have platforms in India and Asia. Why? Because exactly what you just said, under underdeveloped regions, from the interconnection point of view, [00:13:20] they needed to be upgraded infrastructure wise, and we will see this need even more in the future, what we call edge.
It's not enough to serve the hubs. Cities like Frankfurt, New York, la, Miami, Berlin, Bulgaria is a kind of a hub for the Balkans. Really. Interconnection really needs to get deeper into the edge for one very simple reason. Quality security performance in general or a magic world of latency. Latency is extremely important nowadays.
I love to say latency is our new currency. I
[00:13:57] Matt Trifiro: know that was one of the quotes that I saw and I saw in one of your articles that you wrote too, and I think, so let's talk about a, a little lot there to unpack. Okay. So first let's talk a little bit about why is the trend to move the interconnection point closer to the end user of the device and what's driving.
[00:14:15] Ivo Ivanov: One simple answer, performance, quality of experience. If we look into all the applications we all love as private users, but we all need and love and enjoy as business users as well based on cloud technology, cloud compute applications like high def streaming in really good quality online gaming in great experience in.
Real time experience. If we talk about connected cars, the IUTs in general, if we talk about automation, real time manufacturing, automation and robotics, we do talk about extremely latency sensitive [00:15:00] applications and processes. Just to give you perspective applications. All the types I just mentioned, start with really high quality life streaming.
We're not talking about video on demand, we're talking about live streaming, sports events, et cetera, in high dev and even better than high dev, where the experience is great and we all love this and you can't
[00:15:21] Matt Trifiro: easily cash it in advance.
[00:15:23] Ivo Ivanov: Exactly. Yeah, it's live. It's a super bowl. It happens now, right?
Literally this millisecond. And I wanna enjoy it in the best possible quality. And this is exactly what latency is for. So all these applications, they have one thing in common, lower than 35 milliseconds. It's getting lower and lower. It is I in the
[00:15:44] Matt Trifiro: world that I live in, put on my work hat. We're talking tens of microsecond.
For latency. So, yeah, and I think, you know, five years ago that people would've thought that was absurd over the internet. I mean, that's just a ridiculous unit to measure for the internet, which is this best effort network that, you know, kind of cobbled together. But you're right, it's, it's like in order to run these more demanding applications or consume this more demanding content, we have to improve the linkages.
And interconnection is a way to do that. One of the things that so much is curious to me, but one of the things that that you mentioned early on, you said there are two types of networks. There's the content producers and the, and then there's the access networks, the content consumers. I agree with you that for the entire arc of the internet up until recently, that's probably been true, but now we're moving into world.
You mentioned connected cars, where the things in the world are produc. Massive amounts of content, maybe more so than things upstream, right? It's like my car or my jet engine, or my video security cameras generating, you [00:16:40] know, teter bytes data. How is this trend to moving from an internet of humans consuming content to machines, talking to machines, and machines generating kind?
How's that changing interconnection?
[00:16:49] Ivo Ivanov: We call it symmetry of traffic behavior. We saw this especially booming during the lockdown times and the covid, the home. Times, right? Where during home office, a lot of users, associates, right, of companies, they started producing heavy content, having volume because of the video conferencing, right?
[00:17:09] Matt Trifiro: All the schools. All the businesses, everybody's on video.
[00:17:12] Ivo Ivanov: It's smarter than the smartphone. It's actually an ecosystem itself with the thousands of sensors. And I will stay with your question on the connected car example, because it's so interesting. If we're looking to the data related to the connected car, 70% of the data produced in the car are not latency sensitive.
Why They can be uploaded later. Where the car is in the parking space, they're not let is sensitive. Okay, let's look into the remaining 30%. Sure. 20 of those is car to cloud. Five car to car communication. Mm-hmm. . And the remaining 5% of the 30 is car to environment. That means all the sciences on the street and other IOTs in the environment.
These 30%, they have something in. Performance security in the best way possible. They're so latency sensitive. It's about real time communication. It's about traffic lights, It's about avoiding accidents. It's about crucial [00:18:20] performance communication, and this is why interconnection. Needs to get as close as possible and it connection infrastructure to where the cars are, where the users are, or in general where the IUTs are.
This will change the world of how infrastructure is built. We will see interconnection platforms on the, um, highway crossroads. We will see many Dicks exchanges and we'll, we as Dick, were working on this close place close to a 5G cell tower. We are talking about spreading internet infrastructure, including micro data centers with many exchanges in those.
Everywhere, as close as possible where the users are. So
[00:19:05] Matt Trifiro: dks is a platform for interconnection. Mm-hmm. . And when I think of a platform, I think of a universal mechanism, right? So if I adopt a platform, I can do the same thing in. Czechoslovakia that I can do in the United States. When you think of Dex as the platform, is that one of its attributes that you've got a single interface to?
My interconnection,
[00:19:26] Ivo Ivanov: as I said, we, we run platforms, physical platforms. Okay. So physical. It's, uh, very, very secure and, uh, highly performable physical infrastructure behind this. Huge routers or routers at the size of the market requires, Depends on region by region. Um, as we have 40 different, New York is different from Palomo.
Chicago is certainly different from alos in Nigeria, or, uh, Phoenix is different from Mumbai. They all have certain specifics if it comes to market size and grows. [00:20:00] What we have in common is that per market, we do operate a highly distributed. Physical infrastructure platform. That means we do have physical presence in different data centers.
As we're data center mutual, we do not operate data scientist, data center operators, our partners, we also carry jerseys because the carriers isp our customers. So we don't want to compete with our customers. We are entirely as Dick's focused on the interconnection platform. So the platform is built per metro, creating a nice interconnection.
And the different platforms, and now I'm getting to the, to your question or closer to, to what you have, uh, expressed as an assumption that the different platforms, they also interconnected. And of course we, we perform and use a lot of automation these days digits. We do invest millions of dollars for software.
To make it as, as automated as possible for different networks to use it as, as seamless as possible to interconnect to, to, um, to see stats, to optimize their traffic exchange, which makes the platform also very virtual, very automated, let's say smart. So a network operator. Of course can use all the different diggings platforms with the same look and feel in the same way on one portal with APIs, with, uh, a lot of script behind this.
Yes. You know why? Because we are diggings, We believe that interconnection must be democratized. Tell me more about that. Why democratize? Because, yeah, we have been doing this business for 27 years now. 24 years. Of those 20. [00:21:40] We're entirely wholesale industry focused. Just a couple of years back, we started business relationships with enterprises, with car manufacturers, with banks, with retailers, with logistic companies, and they these kinds they businesses, not telecommunications, wholesale.
Their business is the business of their industry segment. They just want to use this type of infrastructure for security and performance reasons, but they don't need to be experts in wholesale telecommunication, so we need to make it as easy as possible for them to enjoy the interconnection service.
This is what are called democratization of infrastructure that every single type of business on this planet can use. Also, internet exchanges and interconnect. In a seamless way without, uh, entertaining hundreds of engineers internally, as in-house staff and, and huge knowledge about this. It's similar to the, in the way what the hyperscalers did for compute.
They democratize compute, and this is, yeah, this is another type of
[00:22:53] Matt Trifiro: I aass instead of infrastructure is a service. This is interconnection to the service, which as you point out is infrastructure. That's really interesting. Okay, so you operate in 40 markets, you said, around the world today. Yep. And in each of those markets you've deployed.
Routers and other networking hardware and then servers to run your APIs and your software and all of that. So you have this physical instructure, but you don't actually own any data centers. You lease from co-location providers, I imagine. Do you have any fiber? Do you lease any fiber or [00:23:20] you just, everybody else has to bring the fiber to you?
We
[00:23:22] Ivo Ivanov: just would just use Doc fiber providers or layer to connectivity service providers, depending on the market, depending on the availability of services on that market. Preferably, we use doc fiber to interconnect the different platforms and the different markets. They're also interconnected mainly on 100 gigabit circuits.
Those circuits we buy and lease. Carriers. Of course this is the way, the way how it works as we are not carrier and is P explained, we are Digg is we're entirely focused on interconnection.
[00:23:52] Matt Trifiro: Well, that's what I was gonna ask you is, you know, you have companies, and I don't need to pick on, but the most obvious Equinix where they have this interconnection fabric and all of this.
And I wonder, I mean it seems like your strategy is very different because when I think of Equinox, I think of like they have these fairly substantial. Points of regional presence. I see what you're doing. Like you mentioned like a local market, like in in Dallas, how many different locations are you in in a metro like Dallas?
[00:24:15] Ivo Ivanov: That's a great question. In New York around 20.
[00:24:18] Matt Trifiro: Okay. And that's very different than these like major interconnection points.
[00:24:23] Ivo Ivanov: You know what the differences, and I have been asked this question a couple of times already. Mm-hmm. , what's the difference between Dex and Phoenix? Because Phoenix has a great platform as well, right?
Mm-hmm. , they have their peering interconnection platform, cloud business exchange for cloud, computer cetera. Similar features as we have as well. The major differences we operate and allow access to our platform in all of the data centers we called enabled, right? Or in very simple words, if you ask me what's the difference between and Dick.
We have more data centers, right?
[00:24:56] Matt Trifiro: mainly be a rack. Yeah. You have more points of presence. I like the more [00:25:00] data centers though. That's very
[00:25:00] Ivo Ivanov: clever hundred to share with you a couple of numbers. We do have more than 600 cities with data centers in those cities, which are Dicks enabled six. Hundred. How are those too?
Right? Because the market, let's say New York is an example, the ecosystem in York physically enabled in 20, directly with Dick's router and in 20 different data centers, as I mentioned initially, we do have a lot of partners also layer to transport networks. We call partners, they extend the reach of the ecosystem, so a smaller network, which is not in New York, present in one of those 20 data scientists can also get access to the Dicks New York ecosystem, a network from Philadelphia or Pennsylvania.
Right. Using the extension of decades through a transport
[00:25:51] Matt Trifiro: network. So, So if I'm Philadelphia and I want to improve my connection to users in New York, one of the ways to do it would be to pay for dark fiber that I control that goes directly into centers and need to take out tendency. That data center in, you know, someplace in Manhattan where you're located and then it's a physical connection I imagine, or.
I could lease. Just simply lease the fiber. And you would provide the termination is what you're saying? Right. And then I'd get access to that ecosystem. Is that what they call remote interconnection? That's remote interconnection. I don't have any equipment in that facility. I'm just using a connection.
Thank you for that. I, I've always wondered what remote interconnection meant.
[00:26:27] Ivo Ivanov: Absolutely. They're different terms describing the same thing. Virtual access, virtual switch, whatever you call it, is actually a remote interconnection. Extending the reach of an e. Yeah,
[00:26:39] Matt Trifiro: because if I was in [00:26:40] Philadelphia prior to D Kicks, I might have to go up through Chicago or Virginia or some other place.
[00:26:46] Ivo Ivanov: I see. And if you, if you don't have the size as a business, you don't have the size, probably you don't want to use the CapEx and opex to be physically present in New York, in Manhattan, in Jersey, or Long Island with your own point of presence in one of those bigger data centers there and enabled data centers for good reasons.
Your reasons, right. In this case, you can still. Use a direct interconnection. I can borrow your presence. That's exactly the point. And latency wise, it's still in the range. Where it's okay. We are wi within the 35 minutes. Yeah.
[00:27:19] Matt Trifiro: For a certain class of applications. Yeah, you're right. That's very tolerable.
Now let's talk about in a metro, right, how do you plan how many spots you're gonna have and where they're going to be and where you're gonna start? Like what drives your analysis of a marketing, your initial footprint? Mm-hmm. , as
[00:27:34] Ivo Ivanov: I said, we wanna make it as easy as possible for different type of network operators to connect to the platform.
So we try to enable as many as possible data centers. Criteria or density of networks criteria are potential for that density and not only on the wholesale side, so network operators like ISPs, carriers on on side and CDs and common networks on the other side, but also what I mentioned just a couple of minutes back, enterprise.
Heavy data scientist. Dick is not a wholesale peering exchange into connection platform only. We do provide different action services, cloud exchange for cloud connectivity. We do have, for instance, um, services like maps, Microsoft Azure appearing service, which is a layer to. Direct connectivity [00:28:20] to Microsoft 365 applications.
The teams, the Microsoft Dynamics, E R P, crm, Outlook, et cetera. Crucial applications for a lot of businesses, enterprises, Organization Day, so they can use directly the Dickus platforms connecting to Microsoft 365. So also nowadays, make our decision. To set up a router also based on enterprise density of customers in the respective data center.
So actually the mix of different type of data centers. You have data centers where more enterprises are in, You have others who are more carrier hotels. So actually the mix makes it beautiful.
[00:29:01] Matt Trifiro: Now in a given market, So imagine you're geographically right, You're spread across. Dallas is a huge market, right?
And you're spread across, you know, probably 50 square miles, I don't know, some large area. The thing that drives you to go into, you know, six or seven or 20 locations in a market is the density of interconnection. Where are the enterprises are and all those factors that you mentioned. Is there also a. An element to like the specific local geography.
And I know that 35 milliseconds, the distance between, you know, the data center, infomart, downtown Dallas and you know, some other data center that's maybe 50 miles north doesn't really matter. But when you start looking at some of these, you know, connected car and real time applications, does the, the fact that you've got many data centers within a local metropolitan region give you certain advantages in this new world where latency.
Privacy and security and guaranteed delivery is maybe more important than other things.
[00:29:53] Ivo Ivanov: This is certainly an advantage, the distribution having more and more data center enabled, but at the same [00:30:00] time, I will not stop there. As I mentioned, our approach is to be present in first all relevant s around the globe within the next three years, but also.
Create different type of platforms from the size point of view for sure, than the one in New York will be smaller, but also to create platforms in tier two TF three, TFO markets getting really deep into the edge because this is what's all about the cars. They drive everywhere. They do not decide on the data center latency point of view should they drive further or not.
At least we don't wanna have this situation. People wanna enjoy unlimited mobility and the internet infrastructure needs to follow and and create the foundation for it.
[00:30:44] Matt Trifiro: Yeah, that's really interesting. Like mobility has certainly made an impact on the way we think of the presence, like when you're delivering content into people's home.
The homes aren't gonna move, so you have a pretty predictable path. But when you're living stuff into people's cars or their mobile phones, they could be anywhere at any time. And you're right, they, I want the. High performance. I don't want my safety to go to go down in my car because I drove another 10 miles past the interconnect
[00:31:07] Ivo Ivanov: point.
Exactly. You're living the secure area of internet, uh, connectivity, performance. Stop driving now, right? .
[00:31:15] Matt Trifiro: Is there a world you think, where you see, you talk about pushing intersection all the way out to the edge, and you mentioned these large companies like banks and stuff. Does that make any sense to actually push your interconnection fabric into the facility that your customer.
[00:31:28] Ivo Ivanov: Probably not a fabric, great equipment, but we certainly, we do have started projects where we will most probably working with partners, of course, on the Esteban layer, et cetera. [00:31:40] We will have a program delivering excess boxes. Into on-prem, access boxes on-prem to customers where we can better cover the last mile, the last mile between their fabric, their manufacturing facility, whatever it is, their manufacturing facility, their campus, whatever it is again, and the access to them, um, interconnection infrastructure.
So what I certainly can imagine is that decades will, uh, deliver access boxes. Enterprise customers, enabling them to connect from different branche. And manufacturing facilities easier to the bigger layer, integration layer of interconnection. So when you, you look
[00:32:26] Matt Trifiro: at this world we're moving into where there's a lot of local interconnection.
You know, one of the things, again, when you started, when you explain interconnection, you explain it historically, which is like, there's content providers and they're central, and then there's users which are everywhere. And so we need to bring in a connection closer to the users. But when we're in this world where the devices are creating a lot of content, like the car.
The 30% that needs to be going somewhere and it probably needs to be processed locally, at least some of it. So you've got your cloud there too, probably. It seems to me that there's a new traffic pattern that might emerge, and maybe you're even seeing this, which is, I guess I'd call it east west instead of north, South, but maybe it's not even east West, but it's like it stays within the region cause it doesn't have to leave.
Are you seeing a trend that
[00:33:06] Ivo Ivanov: way? localizing, as I said, for latency reasons, for security reasons instead of pushing traffic around the globe and let circling, right, this is first from the latency point of view, nightmare. [00:33:20] Second, from security point of view, the shorter the path between the source and the device.
Using or producing traffic, the better, right From security and Poston point of view. So the localization happens, and this is the driver for this edge infrastructure. Let's stay with this with example of what you just mentioned. Regionalization and education of traffic patterns means that we will see more and more concentration of mini t3, t4.
Cloud Edge setups from the hyperscale, they are projects which have been made public already. Right. We do see setups from companies like Facebook, Microsoft, on the content side, not on the cloud side, is only Netflix, Amazon, etcetera. If it comes to different applications, they also offer in not just the compute or the cloud business, they do have.
They have created a, their end extremely flexible and agile edge setups already. So if this is the trend for obvious reasons, what is the missing piece? The Edge interconnection platform aggregating all these different edge devices from Microsoft, Facebook, Netflix, Amazon, etcetera, and interconnecting those with the local fiber guys, with the local ISPs in iva, in Oregon, in South Dakota, North Dakota, everywhere.
Where we do not see today. Big cops as we have those on the east or west coast directly in cities like Miami, New York, la or Seattle or the Bay Area, Dallas is is a hub, but there is so much need to be deployed in the rural
[00:34:59] Matt Trifiro: [00:35:00] areas. Well even not in the rural areas. You mentioned like Phoenix, which is a major US city.
But there isn't much interconnect. There wasn't until you got there. Right? There wasn't much interconnection in Phoenix, not like Dallas or Chicago or Miami or any of these places that it merged historically. That seems inevitable that the trend is going to bring more interconnection to more locations farther out into the edge, into these tier two, tier three, tier four cities.
And I think it's really interesting that DKS has like, I think you're a leader in that. Like you're almost starting. Outward in, you're saying how far can we go and then how can we bring it back into the rest of the internet?
[00:35:35] Ivo Ivanov: I think in the future we need to develop these for the pieces of edge infrastructure, and it's not the dex platform only.
We need to see more edge data centers. Yeah, we need to see more fiber in the rural areas. Middle, middle and last mile investments are extremely crucial also for the economic. Country and serving the people, disregarding where I believe and work with the best possible quality of access to digital services.
Because if you look at from the more altruistic point of view, right, a good access to digital services today means also good quality of life. Because we so much depend on this economic growth at the end, I'm 1000% sure about this. Diversification of infrastructure is extremely important also for redundancy, resilience, security reasons.
Look, we started a market on the East coast, a platform enrichment where I believe. The Richmond has a close proximity to Virginia Beach, where a lot of recently non operational subsea projects, they do want subsea sub cables land, and we [00:36:40] see a new traffic pattern in the MidAtlantic corridor. How interesting.
So we need further the specification of platforms on the East coast because Aspen, New York and Miami, they cannot do the. We need more and more ecosystems, even on the east and the west coast. So it's not about should we have another one or two hubs? No. We need thousands of hubs and, and thousands and hundreds of, not probably everywhere, a real hub, but at
[00:37:05] Matt Trifiro: least Sure.
Some place where networks can come together and, Yeah.
[00:37:09] Ivo Ivanov: And you know why I wanna share this with you. I did not exercise analyzing the behavior of, uh, network, network, corporation, behavior of a companies like Netflix and. Just eight years back, eight to nine years back, they just give these to us example, right?
They didn't have a single own network point of presence globally
[00:37:35] Matt Trifiro: 10 years ago. Really? That's all happened in the last eight, or, Yeah. Ish. Yes. Yeah.
[00:37:39] Ivo Ivanov: They have their enterprise business, like producing great devices and providing great entertainment services. They were using for their traffic distribution and their application content distribution, CDNs, content delivery networks.
Mm-hmm. today, Apple and Netflix. Are still enterprises, but also they are global network providers. They have built their own network, right? Yeah. They have presence in more than hundreds of different data scientists around the globe. They're investing in their own global infrastructure and you know, we will see this [00:38:20] happening with the.
With automotive companies, with the insurance companies, with the retailers, they will become global network operators as well. And this is an opening speech to a huge, exciting decay of new type of industry involvement. We are just at the beginning of a bombastic wave of. Digital infrastructure will involve and what type of new players we will see Appearing good old names with an entirely new business model, creating their own global network.
Wow,
[00:38:56] Matt Trifiro: that's a lot. There's a lot in there. So I actually, I won't mention his name, but somebody who has a lot of experience deep into the internet in Nano and he said to me that his estimate, and nobody really knows the answer of his estimate is 70% or more of the traffic that we think of as being on the internet is actually on private networks
Which is fascinating cuz I still think of, and I think most lay people still think of as the internet, as this, you know, giant public utility that's access to everybody. But what's really happening beneath the scenes is. It's being optimized by a bunch of private businesses based on their particular needs.
[00:39:29] Ivo Ivanov: And this is what we'll see happening more and more What we'll see organizations building their own internet. So
[00:39:35] Matt Trifiro: a cloud provider might say, That's crazy. That's like owning your own data center. Why would you wanna own your own network? We're better at doing that. You should use our network. So what do think all these companies are excited about owning a network?
[00:39:47] Ivo Ivanov: There will be a mix of both, of course, for controllability reasons.
[00:39:52] Matt Trifiro: They may lease all the equipment and just have control of it versus buying it and deploying and operating it. I mean, is dks gonna go into [00:40:00] that business? Are you gonna offer private network services? No, we
[00:40:02] Ivo Ivanov: offer services. We allow organizations already create their own connection, virtual private environment, so
[00:40:10] Matt Trifiro: they have a relationship with the transit providers.
[00:40:12] Ivo Ivanov: Different thing here, it's a different thing. We offer a concept and started offering this two years back and a pretty new. Now we call it closed user group. Closed user group is nothing else than a virtual, private interconnection environment where a bank, where a automotive company can invite all the buyers and suppliers of data, all the providers of data, let's stay with automotive company, can invite all the different companies who wanna deliver data into the.
And use data out of the car or produced by the car to interconnect with the network of the automotive company in one private virtual environment, Let's call it a virtual platform of this automotive company. And now it comes under the security and compliance policy of the automotive company. So they can aggregate many to one under their controlled environment.
Compliance, controlled, security controlled, performance controlled. And what will happen in the end, as I'm not saying that an automotive company now will start and build a. Mobile networks serving their cards with connectivity. They will still continue using Verizon T-Mobile, you know, at and t providing the Es.
IM cards in the cards, but they will let them interconnect with the application providers at the other end. [00:41:40] Under the automotive company's policy because they want to control the data journey, controlling the data journey, this will allow them assets or virtual assets, so more value for their shareholders.
And this is valid for all types of industry. We're just talking about automotive and finance. To examples because they're kind of early adopters. We will see the desire and the demand for higher controllability level of controllability on the data journey level from all different industry sectors as we don't have a company today, which doesn't depend on from.
Digital services, even a company which produces entirely non-digital product. The processes of this company involved in manufacturing and delivery chain rely already on digitalization. So it's actually everywhere.
[00:42:38] Matt Trifiro: Yeah, that's, It's really inter, I'd never heard of private networks being applied in that sense.
So it's really interesting from like the control of the experience sense. But it's also really interesting in the. Like the network of suppliers in a supply chain, right? Cause you're right, your entire logistical network of a fiscal product requires a lot of, lot of input, a lot of like, and, and that's gonna become increasingly real time and you're gonna increasingly wanna analyze it.
That's a whole parallel universe to the
[00:43:02] Ivo Ivanov: internet. Look in the one example we don about creating digital assets and let's say creating a higher value for the investors, right? And the shareholders. It's also. Reputation of a company today. Automotive examples are great because we all use
[00:43:19] Matt Trifiro: [00:43:20] cars, right? And if my internet connection in my Audi isn't very good, I'm gonna blame Audi.
[00:43:24] Ivo Ivanov: Exactly. Imagine if you have paid 150 plus thousand for a luxury car like the equal S of Mercedes, the food electric one. The new S was, and you are sitting in the back and wanna enjoy as a manager. A team's call and this team's call doesn't. What do you think then say, This did a back job, or Microsoft does a bad job.
Of course you'll think about Mercedes and say, This car, Well,
[00:43:51] Matt Trifiro: especially if my neighbor has a BMW and his teams works, awesome. Then I'm definitely blaming. Yeah, that's really, that's,
[00:43:58] Ivo Ivanov: that's really great. Or even, even even the more stupid example, like, we all know this, this feature with the app on, on the smartphone where we, we just click on the button and the car gets unlocked or locked.
If you click on that button and nothing's happened, right? What, What do you, How we feel? We feel bad. Push it again. It's broken. Yeah, you don't think that probably the application BMW, Ford, or Mercedes uses is not that Well interconnected or between the at and t mobile network and the cloud compute zone of Amazon, you don't really
[00:44:36] Matt Trifiro: wanna care.
You don't even wanna care. You just want it to work. Yeah.
[00:44:39] Ivo Ivanov: Oh, this card doesn't do the trick. I don't like it. And this is exactly. What they want, They want to make assured that they do not lose their quality promise. Yeah. So let's
[00:44:51] Matt Trifiro: talk about quality for a minute because, you know, the internet is known as a best effort network, which is great because it's comparatively low cost of, you know, [00:45:00] having a direct line somewhere, but you, like, you throw your, your bites onto the internet and you like hope they get to the place you need to in time.
Uh, and they usually do, but not necessarily. But do you see a world where these big brands need to control that experience where, Every service, interconnection, transit, all these are gonna start offering like guarantees and SLA backed kind of private internet services. Is that, do you see that
[00:45:22] Ivo Ivanov: 1000%? It happens already.
It is happens already. Tell me, can you gimme an example? Dicks offers SLA on the map service. We do offer SLA on the peering already. We do offer SLE on our cloud exchange and cloud route. This is reality
[00:45:39] Matt Trifiro: already. Yeah. Yeah. Cuz you're right. If you're going after that customer that cares about quality and performance, you have to deliver those things.
Absolutely. That really interesting. Just you have an api. Does that mean that I can, for all intents and purposes, programmatically control my interconnection? Does that happen at machine speed? Is this like real time? Can I, It's real
[00:45:56] Ivo Ivanov: time. It's real time, Almost real
[00:45:58] Matt Trifiro: time. And then are you, are you providing me with telemetry so I can measure?
[00:46:03] Ivo Ivanov: Yes and yes. No, our portal is extremely advanced and as I mentioned earlier, we invest really substantial amount of, of money to develop a lot of code here.
[00:46:14] Matt Trifiro: Two years ago when I started the Over the Edge podcast, it was all about edge computing. That's all anybody could talk about. But since then I've realized the Edge is part of a much larger Revolut.
That's why I'm pretty proud to be one of the founding leaders of a nonprofit organization called the Open Grid Alliance for oga. The OGA is all about incorporating the best of edge technologies across the entire spectrum of connectivity. From the centralized data center to the end user devices, the open grid will span the globe.
And it will improve performance and [00:46:40] economics of new services like private 5G and smart retail. If you want to be part of the open grid movement, I suggest you start@opengridalliance.org where you can download the original open grid manifesto and learn about the organization's recent projects and activities.
Including the launch of its first innovation zone in Las Vegas, Nevada. So let's switch some broad, a couple broad topics. Just, you know, from, and you've covered a lot of this, but I'm interested in thinking back, like what, when you look out five years, what is our world gonna be like? How do you see the world beings different in five years?
It
[00:47:12] Ivo Ivanov: will be a world which will use more and more digital services as we can even imagine today. Further down the path we're talking about the metaverse. We're talking about an internet of a new generation or digital services of a new generation where people will enjoy different virtualization.
eCommerce will change, will become extremely al augmented and virtual reality driven mobile. See the business world further. Progress on processing on on different solutions for any type of business that's relying on cloud comput, but also relying on digitalization and digital infrastructure. I see a world full of innovation that see a world full of digital infrastructure.
And demand for digital infrastructure. Therefore, I do worry a bit that we will not be fast enough to develop this digital infrastructure as much as we can and everywhere, literally everywhere around the globe, at least this is our mandate and, and will not stop working hard every day to do
[00:48:15] Matt Trifiro: this. Okay. So that's a great, that's a great vision.
What do you think the [00:48:20] industry could do, Your peers, the governments, all that? Cause you said we might not be able to build it out fast. Well, so first of all, what does that mean?
[00:48:26] Ivo Ivanov: I believe that there will be different beats in development of infrastructure as this is a natural thing. We have seen this in the past 10 years as well.
There are regions with, with less developing infrastructure. I hope we will be able, at least for a solid foundation to close this gap. Probably not every single region would have, will have the high end. But at least a solid foundation for good experience on digital basic services or state of art services should be delivered.
And I see one thing which might happen further, I believe that certain gaps on the infrastructure level, which cannot be closed. By wholesale market participants today, they will be closed by other market participants. And I wanna give the example of what happened with the subsea cable business. The Subsid cable business just five years back, were mainly controlled by incumbents.
The telcos, they invested in subsid cables and. So that they do not invest or they didn't invest that fast, so there was a risk to have a shortage there. Literally within just two years of time, we saw huge projects where Microsoft, I spoke Google and, and further they started investing the millions and millions of dollars, billions of dollars building their own subsid.
Just to make sure that there is no shortage, [00:50:00] and you know what? This will happen also with industry giants like banks, et cetera. It started to happen. There was an announcement for a first subsid cable. By a huge bank from the United States of America across the Atlantic Ocean for latency purposes. So these kind of gaps, and what I predict is that in the future we will see new type of players investing actually in infrastructure built.
We call automotive, we call finance, we call retail. They will, they will get their hands on this as well. The legacy industry stakeholders will not be able to make this investments on time.
[00:50:45] Matt Trifiro: And I think one of the interesting that we've talked about that doesn't get talked about very much is this like the regional and local characteristics for the internet.
This like physical spatial quality to it. And imagine that some. Some locations are more advanced than others. Some cities are creating the demand for these new services and these SLA backed type networks and private networks and all that. Where in the world, like specifically city level are you seeing the most innovation
[00:51:09] Ivo Ivanov: now?
For sure we do see lot in, in the US happening on the the west and the east skills, but also Asia and Europe. One of these three difference. So I'm talking about East Asia, countries like Korea, Japan, a lot of things are happening. China for, for obvious reasons, they have companies from China, they have done also, uh, good progress.
But we do see, fortunately, this happen in, in Europe and still to a very high degree in the US on the east and [00:51:40] west coast, mainly,
[00:51:40] Matt Trifiro: no specific cities that you wanna name.
[00:51:43] Ivo Ivanov: Ah, cities. It's, it's hard to say cities, uh, interesting for innovation because you know, the companies who do. Provide innovation or create innovation today.
They're global operators. They have their different centers. They have different campuses around the globe already, as given one example, Volkswagen, the common manufacturer. They have created their digital innovation excellence centers in Barcelona. Mercedes in Lisson in Portugal, we do see campuses of Apple, Amazon, and Google around the globe where cool things are designed and happened.
So it's really hard to say that's the Silicon Valley. They are many different Silicon Valleys nowadays around the globe. India is an inspiring example. Oh, Dick's Mumbai became number one in Asia within four years of operat. We passed on all the other platform operators in Singapore, Hong Kong, you name it.
Just within four years, the market is like a rocket. It's amazing to see how India grows in terms of digitalization and services, and I predict this will be one of the big, big stars in the future in terms of concentration of innovation and digital infrastructure, but also services. I agree
[00:53:04] Matt Trifiro: with you. I think some of the most amazing innovations are coming out of India.
It there's a lot of really smart, educated, , less expensive, which encourages business doing just amazing in. So Yvo, this has been a terrific conversation. Thank you so much. If people wanna find [00:53:20] out more about dks or connect with you online, how can they do that?
[00:53:23] Ivo Ivanov: Oh, there are so many different options. Just visit the DKS website on the de ci.net.
Feel free to join my LinkedIn profile as well. I try to answer all the questions there, but the Dix website is probably
[00:53:36] Matt Trifiro: the most, Yeah, and I like your blog. Your blog. Nice. So usually read the D kicks blog. So Eva, thank you very much for joining us on over the
[00:53:41] Ivo Ivanov: end. You're very welcome. It has been exciting.
Talking to you and looking forward to catching up in person very soon. That does it for
[00:53:49] Narrator 2: this episode of Over the Edge. If you're enjoying the show, please leave a rating in a review and tell a friend. Over the Edge is made possible through the generous sponsorship of our partners at Dell Technologies.
Simplify your edge so you can generate more value. Learn more by visiting dell.com.