How Equinix wants the internet to be ready for the metaverse

Equinix is one of the companies that provide the backbone of the internet, and it has more than 10,000 companies using its more than 240 datacenters around the world. And so it has a say in how we will build the metaverse, the universe of virtual worlds that are all interconnected, like in novels such as Snow Crash and Ready Player One. 

Intel has said that the making the metaverse work in real time for billions of people will take a thousand times more computing power than we have today. That’s pretty intimidating, but to Equinix, this has been coming for a long time and the latest view of the metaverse is an evolution of the internet. Equinix has doubled the number of datacenters that it had a decade ago.

I talked with Matt George, director of digital transformation and GTM strategy at Equinix, about the technical challenges of building the metaverse. One of the big challenges is not bandwidth, which enables you to download a lot of data, but latency, which determines how fast your interactions can be. The metaverse isn’t really going to be the metaverse if we have high levels of latency. It has to be snappy.

Whether we can do this depends on the datacenters of the world. I’ve been inside one of Equinix’s datacenters, and they’re vast.

Here’s an edited transcript of our interview.

VentureBeat: Tell us about your background and Equinix’s view of the metaverse.

Matt George: I’ve been at Equinix nine years, so I’ve seen the company transform itself quite a bit. I’ve had several roles at Equinix. When I first joined, my background–I’m not a data center guy. I worked in the media industry in the U.K. and Europe for the likes of Sky and BT. I was involved in IPTV launches, and also Tivo on the DVR side. My role at Equinix covers subject matter specialty in content and digital media, but I also manage a team that sits as an overlay between our SaaS function and our technical solution teams. We’re involved in thought leadership. We’re customer-facing. We have a very interesting role within the company.

From an Equinix point of view, when we talk about the metaverse, for us it’s always existing in some aspects. It’s not a new concept. It’s obviously gained a lot of publicity in the last six months, last year, especially because of the presentation by Facebook and their change to Meta. It elevated the subject. But for us, the foundation of Equinix, we see our platform as enabling companies–effectively we have 10,000 companies that are using our platform and building out much more flexible digital infrastructure. That, for me, is the core that will enable companies to thrive in the metaverse.

As you probably know, you can see a number of articles that are very pro-metaverse, and then there are a number of articles that view it as a marketing wrap-up. But coming through that we see the need–our platform has global reach. We have 240-plus data centers. Those data centers are interconnected. That gives any company a platform to be able to cope with the impact of digitization. Whether that is deploying physical points of presence, whether it’s connecting those to edge locations to serve end users with the low latency that’s required, or whether it’s connecting their own ecosystems. Companies that we work with in the gaming and media industry, which are closest to the metaverse, we’re seeing numerous use cases for the platform, and also connecting to other partners.

For me it’s always existed. The basis of the metaverse is that interconnection, that interconnectivity that’s needed to support low latency, to support the high compute that’s needed for some of the virtual uses. That’s where we sit.

VentureBeat: I’ve had good conversations on this topic with companies like Comcast and Subspace. Cable Labs as well. Subspace was interesting because their plan was to try to light up a lot of dark fiber and bypass bottlenecks in the internet that could improve real time communication for games and get rid of more latency. It’s an interesting proposition where they could turn around and promise good connections to the game companies. “If you have a player in Siberia and they want to play a multiplayer game with someone somewhere else, we can make that happen so they don’t have a bad experience.” It was almost fixing the internet one connection at a time. I do know they also had a bunch of layoffs now, though. Maybe that plan isn’t really–it almost seems like a Band-Aid for the metaverse. You’re fixing it for some players, but that’s only one problem.

The interesting thing I learned from Comcast and Cable Labs was they’re talking about prioritizing traffic in the home. If you had a game or streaming video, that might get more priority than something else. Being able to distinguish and identify the traffic in the home would allow you to get better real time performance with less latency. There are these things that people seem to be trying, and I’m curious what floats up to your level around how to make something that wasn’t really designed for real time to work with real time.

George: Companies access the net behind Comcast or of the other network hubs that are all over. Where we sit in this whole equation is at that base layer. A lot of gaming companies have those challenges, the user experience challenges, concurrent usage challenges. A lot of them come to Equinix because they can solve those problems altogether.

We work with a company called i3D.net, a gaming platform. They’re one of the companies that would sit in the metaverse. But their business challenges are exactly what you said. How they work with us is that they can get direct connections to the likes of Comcast and any other networks or ISPs that get them close to the eyeballs. They will use, effectively, our platform as a map of the world and build out locations close to the edge, so that the latency is as low as possible. It’s not only for gameplay. It’s for things like when they want to give software updates or new releases. That’s very similar for a number of other companies we work with. The use cases we’re seeing are very much around user experience, managing concurrent usage, managing peaks and troughs.

But you need this reach and this platform, because–we don’t create services to support the game industry. It’s our base layer of connectivity. There’s a little infographic from Newzoo on the metaverse. It shows you very nicely where companies sit in this sort of relationship. We didn’t speak to them at all, but they put us down at the bottom, which is great. We’re that base layer. There are a lot of companies that will look at services that are supporting this high compute, low latency. Our view is that if you can connect to those within our platform, then it’s building out that ecosystem that makes it better.

VentureBeat: How much capital investment has to happen in order to accommodate the metaverse? I know that some people have put things out there. Intel had said that they expect the metaverse, working in real time with hundreds of millions of people interacting, would require a 1,000 times increase in computing power in the coming years. They put that stake in the ground and said, “You guys have to get real about this.” The spatial internet won’t happen overnight.

George: There’s that balance. We’re talking about a virtual world, and the virtual world won’t happen without physical infrastructure, which is part of the irony here. Where we sit, we see that balance of what we call physical infrastructure and virtual infrastructure, and a couple of other things are driving this.

Typically, 10 years ago, when you talked about an Equinix location, you had to physically own the kit and put the kit in a location and manage it. Now you can spin up services virtually. You don’t even have to set foot inside one of our locations now to connect and start benefiting from this compute power and low latency that’s required. It’s a balance. Companies don’t want to sink huge amounts of cost into physical infrastructure. They want to consume it on demand. You’re going to see companies looking to grow the metaverse looking at these types of opex models, as opposed to sinking capex into further investment and commitment that they may not like.

Certain industries, I think–we’ve talked about gaming. We’ve talked about streaming. For me, again, with the metaverse, it covers a lot of that virtual reality. You look at some of the industries like the automotive industry, the fashion industry–some of those industries are going to look at how they can monetize that type of technology, whether that’s VR or AR or bringing that together. But like anything, they’re not 100 percent sure. They’re dipping their toes in the water, because they don’t want to be left behind.

Again, I go back to the position we find ourselves in. We wouldn’t describe ourselves as a metaverse company, but we’re going to see some business come our way because of the desire of companies to be there. But the virtual world is not going to be enabled without the physical world.

VentureBeat: How much capex do you generally spend in a year?

George: Our capex is built into investing in our data centers, our physical footprint. I’m sure we can send you our latest financial figures in terms of how much investment we’ve had building out those locations. It’s billions, and not just in new locations. Each of our data center locations is a campus. If you were to pick a city, there’s probably about 10 locations connected across those metros, and each of those locations will have a build phase. The demand for space is increasing.

Our business model is split across what we call our hyperscale model, where you have the companies that require huge amounts of space – the Microsofts, the Amazons, the cloud providers – and our retail model, where you have enterprises that are shifting to this off-premise deployment. Those figures are available and we can give you those. That will be a sign of the global demand for both physical footprint and also what we now call digital infrastructure services. That growth will continue.

VentureBeat: I’m interested in this global view of just how much investment is going into the internet, and also how much dark fiber is out there. Is it disappearing? Is it still there?

George: The other thing as well is the growing importance of the sea cables, that network. A lot of those land at Equinix locations as well. You’re getting a very connected reach across the world using those deep sea cables.

VentureBeat: When I talked to Comcast they said that their operating plan was always to invest 30 percent ahead of where the network needed to be, investing for a couple of years down the road beyond what was necessary for that particular year. That’s how they were able to grow 30 percent during the pandemic years. It’s interesting to see what kind of growth can be absorbed. What did the pandemic actually cause to happen?

George: From our point of view, we saw a significant acceleration in the deployment of digital infrastructure. Our business focuses on digital transformation. We were seeing transformation across all sectors pre-pandemic, but the pandemic, as you probably know, just accelerated industries that were maybe behind the curve, and just totally drove industries like streaming, home entertainment, and collaboration tools that were already on an upward curve. The demand for space, the demand for what we classify as interconnection, is enormous.

The other thing that might be of interest to you, if you’re looking at the size of the internet–we measure the private interconnections. We produce something called the Global Interconnection Index, and we’ve been doing that for five years. That tracks the amount of private interconnection that goes across our footprint and others. That’s projected at about a 45 percent growth rate over the next four years. That would be something good to dig into if you’re looking for a general view of how companies are going to connect and the traffic that’s going across those connections.

VentureBeat: For the metaverse, I guess the logical question is, does the internet need a redesign?

George: It depends on whether it’s a redesign or a pivot. If you look at how things are shifting, some of the macro trends, we have more devices connecting to the internet. We’re creating more data than we can properly use. The other thing that isn’t technological is the demographic shift. We have everybody now in buyer mode over the age of 16 who’s a digital native. They’ve grown up with technology. They’ve grown up with the internet. They’ve grown up with this desire to change, to adapt very quickly.

Your question is a good one. As to whether it needs it, I’m not sure if it needs it, but it may be driven. And of course what we haven’t talked about the metaverse is the linking with the finance side, the cryptocurrencies and things like that. I look at it from the media side, because that’s my area, but there’s an area of our business that would have some insight into that financial services side as well. I’m not sure yet. I’m not sure whether it needs it, but I’m not sure whether we’ll have the choice.

VentureBeat: I also wonder whether the metaverse’s demands on the network will be paid for by something else. The example I like to talk about with Nvidia–they’re doing the engineer’s metaverse, the OmniVerse.

George: Yes, Nvidia is a partner of ours. We work closely with them.

VentureBeat: With the OmniVerse, what’s interesting is there are the gamers and game designers who want to build the metaverse, but it requires a tremendous amount of asset creation. The good thing for them is that all these enterprises are building those assets for things like BMW’s digital-twin factory. Once BMW builds this, it becomes a bunch of assets available in the OmniVerse. They’re interoperable. And then Nvidia was talking about how they may have to build out all of North America in order to test their self-driving cars, because they can’t be tested in the real world yet without killing someone. They’ll test them in a virtual world, and that has to be a digital twin of the real one. They need to build it out for that purpose, but again, game developers could use those assets.

Nvidia’s most ambitious thing is Earth-2, the digital twin of the Earth. They want that to be accurate on about a meter-level scale, because then they can feed all that data into supercomputers and come up with climate change models, predicting climate change for decades to come. That’s their most ambitious product. I asked their CEO about that and said, “Well, if you’re going to do that, don’t you get the metaverse for free?” And he says, “Yes, we get the metaverse for free.” Is the demand for the metaverse, then, going to be paid for by something else that’s already happening?

George: Nvidia’s a great example to talk about, because they have use cases on the immediate side. They’re working with a lot of the car companies. They’re working with BMW. So yes, as they develop different things, different models will be monetized to enable them to have a business impact. What they’re looking for, really, is this end-to-end collaboration. It goes back to where–we have always talked about an ecosystem. An ecosystem is really a smaller version of a metaverse. Every industry has an ecosystem. If they collaborate, if more businesses now will work on this on-demand, needs-must basis, then things will scale.

You’re going to see certain industries, certain companies, adapting very well. You’ll see certain other industries maybe at the moment that are just trying to work out what is and isn’t relevant to them. The next thing you’ll probably get is the metaverse’s relationship to sustainability. Is that good or bad? Those are some of my thoughts on it. But where we sit as Equinix, we’ve seen some of this happen before. It might not have been called the metaverse. But it’s the foundation of our platform. It’s going to be fueled by collaboration and interconnection.

VentureBeat: It sounds like you would probably fit in the category of a close follower of the internet, and a metaverse optimist?

George: Me personally? I think so. Again, I wouldn’t classify everything I’m talking about as related to the metaverse. For me it’s something that is already happening. You don’t necessarily need to put a label on it. But the Nvidia stuff we’re close to, because Nvidia is a partner of ours. To enable some of the things that they want to do, they will align with Equinix. Jensen Huang talks about the Equinix relationship in some of the YouTube videos that are out there.