The Data Center Outlook for the European Market


In this episode, we talk with Head of EMEA Data Centre Solutions, Advisory & Transaction Services for CBRE, Andrew Jay, who walks us through the future of the European data center market.

El contenido de los recursos se muestra sólo en inglés.

Announcer: Welcome to Not Your Father’s Data Center Podcast, brought to you by Compass Data Centers, we build for what’s next. Now here’s your host, Raymond Hawkins.

Raymond Hawkins: Welcome again, we are recording today on September 16th, here in 2020, as the world continues to wrestle with the global pandemic. And today on Not Your Father’s Data Center, we are joined by Andrew Jay, Head of Data Center Solutions for CBRE, for the Europe, Middle East, and Africa. Andrew, thank you for joining us.

Andrew Jay: Pleasure, nice to be here Raymond.

Raymond Hawkins: Andrew, we’ve had the good fortune of having CBRE on before to talk about markets. And the last time we did it, we did it with Hayne Schrader, and talked about two or three key North American markets, would love to do the same with you in Europe. And if you want to expand the aperture, and talk a little Middle East, or a little Africa, feel free to do that.

Raymond Hawkins: But we’d love to talk a little bit about what you see in the marketplace in London, how the world is handling the pandemic, and how it’s impacting the data center business, and then any other markets that you’d love to highlight, and talk through, we’d love to hear from you and CBREs expertise.

Andrew Jay: Yeah, that’s great. Do you want me to just launch into it Raymond?

Raymond Hawkins: Yes, sir. You can fire away Andrew.

Andrew Jay: It’s been an amazing journey really, since we saw the massive uptake in demand for data centers across Europe, and we look at the big market. It’s been London, Frankfurt, Amsterdam, and Paris, that’s in sort of descending size order. London’s that the biggest market, it’s around 700 to 750 megawatts of colocation space, which is the metric that we use to look at the market.

Andrew Jay: And what we’ve found since 2016, is that the annual take-up has more than doubled. And that’s just astonishing any asset class to have twice as much take up, year on year for four years as you’ve had ever before in history. And that has generated some really interesting dynamics in the market. And what we’re finding at the moment is there’s some real sweet spots particularly in London, and this goes for most of the markets, the big markets, a real sweet spot into the geographic areas where people want to be.

Andrew Jay: And of course, that’s because it’s mostly hyperscale or cloud-driven. With the big three, being the Microsoft, the Amazon, and the Google companies. And they are responsible for taking up something ridiculous, like 80% of all of the colo space in the main markets, which is extraordinary. Everything kind of revolves around them at the moment, and because they’re quite locationally sensitive.

Andrew Jay: And so in London terms, the real sweet spot area at the moment is … I appreciate not everyone’s going to know who these places, but basically out G West of Central London is a place called Slough. And so you go from sort of that far western side, you come into Stockley Park, which is near Heathrow Airport, which most people will be familiar with.

Andrew Jay: And then you come into a place called Haze, and then you go to a little bit further into Port Royal, is a real sort of corridor there on that western side. And we are just having an absolute feeding frenzy at the moment, of developer operators trying to secure the land. And of course, it’s the usual issues around power planning, fiber, and risks. And the prices of these sites are actually going through the roof at the moment.

Andrew Jay: And we are starting to see such competition for the sites that come on, and can be delivered in a reasonable timescale, let’s say, where you can get space up and running in 18 months, maybe up to two years, but not more than that. There’s such intense competition, but now people are starting to say, «Goodness, the cost of the sites needs to be absorbed in some way,» where in years gone by no one really backed, they annihilated the outlaid price of the land. And so now they are building up and building really big. It’s almost got a set of American feel to it, maxing out these sites.

Raymond Hawkins: Andrew, give me some feel, you said that the cost per land has gotten really out of hand, just relative what’s the number?

Andrew Jay: I was really hoping you wouldn’t ask me that question.

Raymond Hawkins: It’s okay. Because we know the Virginia, we are seeing deals, well, million and a half, $2 million on an acre number. And just wondering if that Slough through into Heathrow, that’s chump change compared to London?

Andrew Jay: If you could give me a few acres at that price, I would buy it out of my own pocket almost.

Raymond Hawkins: Yeah, that’s what I figured.

Andrew Jay: Where we’ve seen transactions at 6 million pounds per acre, and I am looking at some deals at the moment where it’s quite considerably north of that actually. And that hopefully just gives you a feel of where it’s gone to and why. Therefore, the operators are saying, «Well, actually, we’re going to have to build up.»

Andrew Jay: And in some of the case they’re looking at four, five, six stories, which certainly in my experience, we’ve only seen in places like when I’d been to Tokyo, for example, it’s big multi-story. Singapore, some of the core, core areas where the real internet hubs are, and there’s like restrictions on land, you see buildings go up. So in London must be the [documents 00:06:23] are, but nowhere else have I really seen that, it’s extraordinary.

Raymond Hawkins: Yeah, 6 million and acre, that’s real money.

Andrew Jay: It is, and before everybody gets excited about that, that is as I say, a very, very specific area. And of course, everybody gets that it’s expensive now. And so a lot of the end-users are trying to then leap frog away from those areas, and get somewhere where they can get a large amount of land at a lower price, less competition for the power. And I guess, it’s all the common sense things that they’re doing.

Raymond Hawkins: And Andrew, we probably should have done this at the beginning, but as you talk about London, you talk about it with a wealth of experience, if I remember correctly this is 20 plus years for you in the real estate market in London, is that correct?

Andrew Jay: Yeah, I’m bizarrely, I’ve been doing data centers for 20 years, but before that I was an investment guy, at City of London, office agency guy. And I wouldn’t have believed that I frankly, would have made a career of advising on data center, but it just shows how the demand for data centers, and the whole world, that ecosystem has just grown beyond all belief in that 20 years sort of, boom, and bust time. Yeah, it’s extraordinary.

Raymond Hawkins: I know Slough, and in around there as the prices are through the roof. And you mentioned that providers are looking at other areas of town, which direction have folks gone? North, where, what’s the next thing as London and in the city gets so expensive?

Andrew Jay: We used to talk about London being a couple of soft markets, which were the City of London, so the financial district, plus the London Docklands, which is where the original data centers were built. And the Docklands is very much the connectivity hub. That’s where people like Telehouse, and the other big companies are.

Andrew Jay: And then we talked about outer London, being everything within a sort of 35 mile radius of that, so that’s a low latency area. And Slough, really started developing properly about eight or 10 years ago, maybe a little bit longer actually, when some of the big corporates went there. So for their financial services data centers, and it just grew from there.

Andrew Jay: And so we now are looking at very, very specific areas, for example, one of the hyperscalers is looking at North London. And so it’s a very particular location, that they’re looking at. And suddenly we’ve got a lot of interest now in land in that area, because where the parent, if you like of the availability zone has gone into North London, then the child sites will be close by.

Andrew Jay: And so what we haven’t seen is say, East London yet, and we’ve got a couple of really good quality data center builds in East London, but the demand hasn’t quite got there yet. I mean, there’s no question it will get there, but it’s not there at the moment. And as they say so much of business life is about getting your timing, right?

Raymond Hawkins: Yes. Is east meaning East of Slough? Further East of there?

Andrew Jay: Yeah, east going back into Central London, and then the Eastern side of central London. That would be quite a way from Slough, the other side of town ready.

Raymond Hawkins: Gotcha.

Andrew Jay: And then the last, we talked about the hyperscalers, but of course, London was built on enterprise demand, and there is still good enterprise demand. If you look at it in the context of the last 10, 15 years of demand into it. The demand is there, but it just gets dwarfed by the hyperscale transactions that are being done.

Andrew Jay: The enterprise it’s interesting, we were worried going into the COVID situation, that enterprise amount we’re going to stop. And in actual fact, all the projects we have on continued and they’re all getting completed. I think the thing that’s changed is we’re not seeing the new enterprise projects come up, and that’s probably because they are just trying to keep the lights on at the moment.

Andrew Jay: They are, I think, accelerating their thought process around going to more of a hybrid, IT architecture, more cloud adoption. They’re looking at cost saving, but they’re not quite there yet, but certainly my expectation is next year in 21, we’re going to see those enterprise companies come back. But with clear strategies around cloud and hybrid, and also trying to reduce down that dependency on the what are now relatively old launch, generally inefficient under utilized corporate data centers, that a lot of them have ended up being assembled with.

Raymond Hawkins: Andrew, we have listeners in all the major markets, and I’d love if with your expertise, and time, engraved there in London and CBRE, do you mind giving us two or three of when an enterprise client comes to CBRE, who are two or three of the best providers there in town in London?

Andrew Jay: Wow, that’s another question or wish you hadn’t asked me.

Raymond Hawkins: 6 million, we already got one tough answer, 6 million acre, or 6 billion pounds an acre, not just dollars.

Andrew Jay: How to win friends and influence people.

Raymond Hawkins: Right, right, right.

Andrew Jay: Oh, my goodness. We are very lucky enough.

Raymond Hawkins: Where the people you mentioned will be happy, let’s look at it that way.

Andrew Jay: Absolutely. We’re very lucky in London, because we’re the biggest market. We have got pretty much all of the main providers. From the wholesalers, to digital realty, to aquaponics, those companies are very, very strong. So the next generation down on the wholesale side, like the Cyrus ones with very, very strong, got several buildings in Slough, very, very successful

And they built up to the pan European footprint. And what we’ve also got is some very good local providers, and unusually for Europe, we’ve got companies like Virtus and Arc, who are both just concentrating on the UK and principally London. What we have found in recent years is as soon as someone’s built up a platform, then they get gobbled up by either one of the big trade players or private equity wanting to build up a new platform and expand it out.

But the likes of Virtus and Arc, have been very, very successful in London. And I think they have significant expansion plans. That and I’m sure I’ve missed people off that that less for which I apologize, and most of the people now, I have to say are good quality. They’re not competing as much like they used to in the old days on the widgets, that got in their data centers.

I think the specifications now are all extremely good, security and operations, pretty good. And so it’s more now competing on where are you? Are you in the right location? And what is the cage system that you’ve got, because we’ve built up different ecosystems now, really distinct ones. And we’ve got the connectivity ecosystem in Docklands, which I mentioned, we’ve got financial services ecosystems in Slough, we’ve got the hyperscale ecosystems, et cetera.

It really depends what you’re wanting to achieve. And there’s very few people, they come to us and just say, «Look, we just want some data center space.» But those that do, then my goodness, there’s a fantastic choice.

Raymond Hawkins: All right, well, I’ve asked two questions that you would have liked to have avoided. I’m going to present this next one as two questions, let you pick which one you want to talk about. I know we’ve only got a few more minutes of our time with you, Andrew. Do you want to talk a bit more about London, in the context of Brexit, and the continent, and how it’s going to impact, or continue to impact data center, or would you rather pick another market in Europe and talk about that in our last few minutes together?

Andrew Jay: I think I’ll take the really easy option there and go for the some of the rest of Europe and talk about that. And I think we often get very centered on those FLAP cities that we talked about, and there’s good reason because of their size. But we are seeing, I mean, an example would be Milan, for example. Milan, I went there numerous times in boom.

And then, I probably didn’t go there for gosh, 15 years. I mean, it’s extraordinary. There was just no demand. It was probably a 15 megawatt colocation market. That’s all is a handful of buildings. And then suddenly, probably three years ago, maybe four tops, we got phone calls from people saying, «Can we have a land in Milan?» And we said, «Well, what on earth do you want that for?» And they said, «Well, I can’t tell you.»

And then we get another call from someone saying, «I can’t tell you,» and suddenly you work out. Well, it clearly there’s one of the big guys is willing to go with there. And so that market has gone from being a 15 megawatt market to SUPERNAP data for a local company, Aruba, have all bought land, built pretty significant facilities.

Currently, that’s probably a 60 to 65 megawatt market. And that has got just with the schemes that are there today that could expand to double that. And probably more, so you could probably go into 175 megawatts now. That’s a 15 megawatt market where we’ve now got line of sight to 175. Now that’s extraordinary and there’s some amazing opportunities there.

And Milan isn’t unique and places like Madrid is similar. And you can literally tell as the hyperscalers start having these secret conversations saying, «Don’t tell anyone, but we want to go to Eastern Europe.» Suddenly, the phones start ringing and these small markets go through a huge acceleration. And it does cause some real problems, because you … And we talked about land prices at the start.

But if you suddenly go into say Madrid, and six people are all wanting to buy a piece of prime data center land, then suddenly, guess what? The price goes through the roof. And we’re finding this in places like Warsaw, we told you about Milan and Madrid, some of the Swedish countries, and Swiss country, so Geneva and Zurich prices are going up, and there’s very little availability.

And it’s also giving a lot of stress to the grid infrastructure, because the grid infrastructure generally in Europe is older and probably not particularly flexible relative to what you might be used to in the states, where you’ve got a hell of a land as well. And so, that has caused enormous problems, plus all of the issues around permitting local code, local language, it makes the whole development process quite challenging.

We were on a call literally yesterday on the Madrid, and the guys were saying, «Yeah, do you know, you’ve got this great piece of land, but the power might take three years to get if you’re lucky.» And people just won’t wait that length of time. And these secondary markets are really interesting, they’re growing. If you’ve got people listening, that want to go there, then they’re great opportunities, but it is quite difficult to get a foothold.

Raymond Hawkins: That’s amazing Andrew, that you could see Milan, go from sleepy for a decade at 10 or 15 megawatts to, it could be as much as 150, 175 megawatts. I think it speaks to how rapidly our world is transforming. I heard somebody say that COVID has caused three years of IT acceleration to happen in six months, I think that’s without a doubt, an accurate statement.

Raymond Hawkins: And I think we are not drawing a direct correlation to the activity in Milan to COVID, but certainly in the acceleration of the desire to operate at a distance, and be able to telecommute, and the advent of Zoom, and all of those things that we’ve seen lots of increased demand on the network as a result of this pandemic.

Andrew Jay: It has been amazing, isn’t it? I mean, we’ve not really talked about the pandemic, but I think that the highlights for me are, we’ve had quite a few construction delays which have been frustrating, but I’ve been quite proud actually of the industry, whereby people saw that the end users, generally the hyperscalers and the developer operators, haven’t very much partnered together to find solutions.

Andrew Jay: And there’s been very much a safety first approach, which has been great in our … I was worried that people would be making the lawyers rich, but that hasn’t happened. And we’ve also had problems with the planning system, and the public utilities, because they are both public processes than they’ve been delayed significantly. And so that’s not helped on this availability point of land for the data center development.

Andrew Jay: I think also the investment market has suffered a little bit. Simply we have a lot of assets, I say a lot, probably three or four specific investments that we were literally about to take to market. And we had to pull those, because investors simply couldn’t get to see them, but I think that’s starting to ease up now. And people are starting to see light at the end of the tunnel, and be able to jump by aircraft, and things to see these assets. I don’t think COVID has had a massive negative impact, but there’s certainly been some time delays, which haven’t helped us.

Raymond Hawkins: Well, Andrew, thank you for the details on London. Are there other markets in your experience that you’d like to highlight for us?

Andrew Jay: Yeah, there are a big market, I guess, from London being the biggest, then you’ve got Frankfurt, and then Amsterdam, and Paris. But just put that into some context, Paris is about 210 megawatts of colocation supply, Frankfurt just over 400 compared to London, which is just over 700. All of these comments need to put into context when comparing it to the likes of Northern Virginia.

But the FLAP markets as we call them, Frankfurt and Amsterdam, and Paris, they’re all just sub-markets if you like to. London, similar things are happening, we’re being driven by the hyperscalers, and I think the really interesting story is what’s happening outside of those core markets.

And we’re seeing a huge amount of activity now in places like Italy, so in Milan, places like Madrid, we’re seeing deals in Switzerland at the moment, both in Zurich and Geneva. And then also some of the Eastern European countries starting to open up, and we’ve seen some transactions in Warsaw recently. There’s an awful lot going on, that is outside of just those main FLAP markets.

Raymond Hawkins: Andrew, just to clarify, you gave 700 megawatts as a London figure. Do you mind giving the Paris, Amsterdam, and Frankfurt again? I just want to make sure our listeners have scale of what FLAP looks like.

Andrew Jay: FLAP’s about 1700 megawatts of colo supply, and the breakdown is, London at 700, Frankfurt for 25, Amsterdam at 400, and Paris 210.

Raymond Hawkins: And can I ask a question, because when I think of Germany, and this might be really stupid sounding. But when I think of Germany, I think of Berlin, I think Berlin is the largest city. How is it that that Frankfurt got to be such an incredible market?

Andrew Jay: It’s a really good question. Germany often it is the country that sort of breaks the rule of the fact that it all happens in that capital city. And that the federalized nature of Germany in part explains how that’s happened, but why Frankfort being so dominant? I mean, it’s just very, very simple going back to boom days.

All the fiber companies felt that people buying their services were going to be financial services companies principally, and they’re all headquartered in Frankfurt. They didn’t perceive that the big manufacturing companies, the car companies, et cetera, were going to be their primary targets. And so they’re all descended on Frankfurt, and they stayed there, and built the eco system, which of course is so important.

And very much London and Frankfurt grew up in very much the same time. Now, the secondary markets in Germany did evolve a little bit in boom. The likes of we used to do Berlin, Munich, Hamburg, Dusseldorf, those types of locations. But then crash came, the fiber builds stopped. And this is why places like Madrid didn’t really take off, because it was all poised ready, and then the builds just were stopped, and it was a like that in the secondary German cities.

But what we’re seeing now is, there’s a sort of push and a pull factor on the German cities. The push factor away from Frankfurt is that, the land prices and land availability, the availability of power married up with the land has become incredibly difficult. And a lot of the developer operators are struggling to find land that works there. And so there’s a sort of push factor away from Frankfurt.

And then there’s the pull factor towards, and you mentioned Berlin, it’s a brilliant example. There’s a pull factor towards places like Berlin, because they’ve got a lot of government functions there, they’ve got a lot of just single large corporates that are there, or they don’t have the financial ecosystem of Frankfurt. And so that’s what’s now starting to evolve and just others are bizarre situation, I had a few months ago, where we were asked if we had any land for data center development in Torin, which is-

Raymond Hawkins: Torin is tiny.

Andrew Jay: Exactly, it’s a tiny, tiny place in Italy. And we said, «We’re not flipping out, why, Torin?» And so we can’t tell you, it’s all confidential. And I mean, I think my theory on it, and of course it’s never been stated clearly. My theory on it is that one massive, great big company that’s headquartered in Torin, and they were looking at some sort of a cloud type outsource deal for which the cloud providers would have needed to provide a new data center.

Raymond Hawkins: Ah, that’s right.

Andrew Jay: That warm really big corporate outsource was enough to drive demand for a tiny city in Italy. And I suspect the secondary cities in Germany is a little bit like that would be in places like Stuttgart had got a handful of very, very large companies like the car manufacturers.

Raymond Hawkins: Right, right, now, I see it. We see that a little bit here in the US too. A broker walk on Leicester City and say, «We need …» And you’re looking at a map and you’re like, «Okay, there’s only two companies there, which one of them is it?» Just like you’re describing into Torin.

Raymond Hawkins: Fascinating. I like your conversation, some push that Frankfurt land has gotten expensive and that it’s tougher to make the economics work there. And some pull that places like Berlin and Stuttgart have enough activity now that they might warrant a further development. That’s an interesting thought of how the demand is getting impacted on both sides of the equation, pretty fascinating stuff. But still 425 megawatts, that’s a huge market. I mean, that’s a lot of capacity.

Andrew Jay: Yeah. It’s a reasonable market, in a European context obviously, it is really big, but we’ve got some markets that are, I would call them sort of secondary markets, which are only sort of 50 megawatts. And a good example would be somewhere like Madrid, and don’t quote now 50 for Madrid, I didn’t have the numbers to hand, but something like that.

And then suddenly you get hyperscalers arriving, and people are buying land, and they’re looking at a development pipeline of 150 megawatts of brand new space. So suddenly you go from 50 to probably 200 megawatts in a very short space of time. And that really does test the power infrastructure, and some quite fundamental issues around just getting those types of facilities built in those size cities.

Raymond Hawkins: Yeah, the hyperscalers, when they show up, they can fundamentally change the dynamics of a market. I mean, especially if you’re talking about a 50 megawatt market, so often we see those guys show up and they want a minimum of 60, or 120, or 240, they want line of sight and incredible amount of growth. And they can fundamentally, if they show up in a city that doesn’t already have that, they can change the city dramatically.

FLAP, I think lots of people from here are FLAP, just to clarify for our listeners, Frankfurt, London, Amsterdam, Paris. I think I understand London, I think I understand the demand behind Frankfurt and Paris. Could you take a few minutes and give us your insights around Amsterdam before we get out and talk about some of the secondary markets?

Andrew Jay: Yeah, Amsterdam doesn’t have big corporates headquartered there particularly, and Amsterdam is very much a strategic connectivity hub for mainland Europe. And again, going back boom, Amsterdam was viewed as a location, which sat at a very stable, very neutral, and it sat between several very, very large countries.

And so a lot of the fiber went into Amsterdam. And if you could just look on the map, you kind of get what I’m saying. And so over that period then, it became very much an interconnect ecosystem. They have the M6 exchange there, I think pretty much every single carrier was in Amsterdam. And it was really for that reason that it grew into what it is today.

And of course now you’ve got, interestingly, I can’t say the names, but you’ve got warm certainly hyperscaler who’s in there in a big way. They’ve taken that connectivity and thought, «Yeah, we want a slice of that.» And interestingly resolve that, because they want to go where the actual end customers are, and have networked themselves in a slightly different way, so they don’t need the connectivity there.

Raymond Hawkins: How do cable landing stations and the cities where those happen … In North America, I think our listeners and we’re pretty familiar with, hey, Northern Virginia, clearly a big spot the West Coast, Northern California, Hillsborough. Where are those hot spots when we think about the continent?

Andrew Jay: It’s a really interesting question. And it’s one that I think I might give a very controversial answer to.

Raymond Hawkins: Oh, we like controversy, it’s good.

Andrew Jay: Wow, so I guess, not controversy. Probably shows my ignorance on the subject, but in-

Raymond Hawkins: Andrew, is in favor of higher taxes and cable landing stations.

Andrew Jay: In Europe, I don’t see a huge amount of demand being generated by cable landing stations, all be will probably … I’m probably really got a statement I can’t say example where it has that, which is Marseille in Southern France. And there are a lot of cables that come from Africa, particularly Northern Africa, that come and land into Marseille. And that’s that touch point for mainland Europe?

And so you’ve got people interaction or now digital realty that in a pretty big way, you’ve got some of the hyperscalers there, and you can actually see that. But with that section of that market, I can’t think of another location other than literally just random ones like Northern Holland, where again, there’s what I think is public knowledge that Google have built a data center up there. They did have a couple of data centers at one point.

But there data cable came in there years ago, and that was a sensible place for them to go, but that hasn’t created an eco system around that at all. Right now, we don’t see the impact, and in fact, I’ve got a lot of examples where in fact, the presence of the cable, people think it’s going to make huge differences to the demand, and it just hasn’t happened.

I think Southern Ireland in Cork, would be a good example of that. Northern Ireland with Project Kelvin, London Dairy, people have built data centers there, and there just hasn’t been a huge amount of demand. So, no, I think it’s something pretty complex, and it’s not as simple as saying, «Build and they’ll come.»

Raymond Hawkins: Not as controversial as you might think. I was hoping we were going to get us a real fisticuffs going on, something … But no, that’s good. Yeah, I think that you give great examples, right? Cork has not turned out to be a data center market, Dublin is, but Cork not. And I think your Marseille example is a good one, I think it’s probably the best example on the continent, where you got an ecosystem around a landing station, but there’s not many others. I would agree.

Andrew Jay: I’m glad that you agree, because I’m sure there is someone listening to this discussion, what on earth is he talking about?

Raymond Hawkins: Yeah, when they email me, I’ll forward it to you, Andrew. So all good, all good. Okay, well, thank you for handling the cable landing station conversation. FLAP, I feel like we’ve got our arms around as a listener base, and as you and I both servicing this industry. I don’t think we have a very good handle on … And you named four off the top of your head, and I loved it.

Raymond Hawkins: Did a few minutes on each one of these secondary markets, and you listed Milan, Madrid, Zurich, and Geneva, but if you have other ones that you find more interesting, I will tell you at compass, we here we’ve had some requests around both Milan and Madrid. We have not around Zurich and Geneva. I know you’re much closer and have a better view on those secondary markets. We’d love to hear about those, and what you think is the advantages, and what might be driving interest there?

Andrew Jay: I think take Milan and Madrid, I think, they are very straight forward, which is, they’ve been pretty small markets particularly now. I mean, Milan was a tiny, tiny market. I don’t think we had a requirement there for 10 years frankly, around time little pops. And Madrid was a little bit bigger, never had places of interaction, and one or two others there.

Andrew Jay: And suddenly it is the hyperscalers who have turned up and that’s what’s just change the landscape altogether. And that’s driven by a few things, so it’s the corporate demand in these countries, Milan and Madrid, they do have some pretty big corporates who are going on that migration to cloud and hybrid IT journey. So there’s a direct business reason for the Cloud people to get there, just look at GDP and population metrics as two simple ones.

That explains a lot of it. And so they’ve descended and suddenly, we go from touched on earlier sort of 50 to 150, 200 megawatts. And just I was chatting to DATA4, who have just, I think, they’ve just finished their building in Madrid. And it’s fully pre-let to a hyperscaler. That’s great, and similar things are happening in Milan.

I think Milan is quite interesting, because that’s where SUPERNAP are. And then there’s again, DATA4, and a company called Aruba there, and they’ve all been fairly successful, and having been successful, you’ve now got quite a few other companies looking at those locations as well. Again, just all hyperscale driven, but people talking about again, tens and tens of megawatts.

It’s interesting your comment earlier, actually, you said that actually you guys want to ramp up to … I can’t remember what you said, but in idea it was tens and tens of megawatts. I mean, Europe’s all together and just a little bit small, you can kind of take a decimal place off. And generally we’re seeing now the hyperscalers saying, «Look, we’ll commit to eight, 10, 12 megawatts, but we need to ramp up to 20 plus.» That’s the sort of scale we’re looking at.

And as a result, people are looking to a minimum of 20 megawatts IT loads for any new building that they build pretty much anywhere. And if they can get three loads of 20 on a modular basis, then that the sorts of sizes that people are looking at, but it soon mounts up.

Raymond Hawkins: Yeah, no question. I would agree with you that my numbers were North American numbers, but we see that the questions we get asked about the markets we’re thinking of entering over there in that tens, and get me to 20, and get me to 30, those kinds of conversations. Well, we’ve done a pretty good run around Europe, and Andrew, I appreciate that.

Would you be willing or comfortable to give us a few minutes on the Middle East and then a little bit in Africa. When I think about where the global population is headed, and I might be thinking of, tectonic plate shifts, decades generations, I think we’ll see more population in Asia, and in Africa, but I’d love your thoughts on either the Middle East or Africa activity.

Andrew Jay: Very happy to talk about them again, relatively short conversation. If we take the Middle East first, we always question whether there’s going to be good demand there. And so I actually went over to Dubai and Abu Dhabi a few years ago on behalf of a client of ours. And the conclusion that I came to was, it was a very, very small market. So any one of those markets in its own right Abu Dhabi, or Dubai, or Jeddah, Qatar they’re small.

And so if you’re going to do anything, you need to be really aggregating some of that demand up to make it into a decent size, to make the economies of scale work for you, to make it worthwhile focusing on deploying capital and time and effort in that very, very small area. As opposed to just going to your next city in Europe, or Asia, or the US, but the conclusion was, it’s just really, really small, and a lot of people they just put there what they absolutely have to have to service local demand, because most global companies they’ll have US, Europe, and Asia, or with heavy compute and storage.

And if you think geographically where the Middle East sits, it’s kind of between Europe and Asia, and it’s like a stopping off point, they don’t need to have big infrastructure there, they can service it from the other locations. So again, a real generalization, but that was what I found. And I don’t think I’ve seen anything that would disprove that sort of theory.

I suppose the watch word on all that of course is, it only needs a hyperscaler to come in and say, «I want 50 megawatts and then everything changes,» but I don’t think we’ll see that in the near term. And then, I mean, Africa for me is gosh, I mean an amazing place, obviously, Northern Africa, Southern Africa, they’re all very, very different, but in terms of probably the biggest market being South Africa split between principally Johannesburg and then to a lesser extent, Cape town.

Yet huge amount of people reasonably commerce it’s on the Southern African locations. That’s the most developed, the highest GDP, most prosperous, stable, et cetera. There is good activity, but again, it’s measured in a few tens of megawatts. It’s not more than that, but there is some activity there, in South Africa at the moment, there are hyperscalers running around and the associated developer operators looking to buy land.

But I don’t think there’s any really big builds going on there yet, but that could be the first one. And then I think there are people with business plans around covering Africa. But those business plans, the ones I’ve seen are all very, very small. Let’s put half a megawatt in Kenya, people like Guy Wilner, he’s always looking at things like that, he’s a great operator. And he’ll go into these little niche markets, but no one’s talking about putting five and 10 megawatts in those top types of locations. So I think the watch word is scale and just scaling your ambitions to the local market.

Raymond Hawkins: You talked about North Africa and South Africa both, I think when I think of Sub-Saharan kind of in the middle, the only one that makes sense to me from a GDP perspective is Nigeria, any meaningful conversation there?

Andrew Jay: Yeah, I’d put it in exactly the same bucket as Kenya. So yes, similar sort of thing there just isn’t a lot happening there, we’ve done a couple of searches recently from I think, Nigeria, Kenya, but nothing’s come of them. But yeah, some something will happen there, but it’s not going to be big scale. If you want to be small niche, Agile, hot and fairly high risk, arguably. But those are probably really good untapped markets, but you’re not going to be building 20, 30, 40 megawatts of IT space over there.

Raymond Hawkins: I’m going to ask you a question and you’re welcome to punt on it if you’d like Andrew. I’ve been to Africa many, many times. My trips have all been faith-based to do missionary work and I was struck when I went into some extremely remote places that no roads for of miles, and I would get there, and people had cell phones, and I was struck by that. I was like, «How is it that you have cell phones?» And they said, «Look, we just skipped over, we never did the landline build out where we had phones that sat on the wall.»

And so when it just became cost-effective to have a cell tower and people would literally in a town where there was no power generation. They had car batteries with adapters, and they use that to charge their cell phones. I use that example to say they skipped over the landline phase, and I’m going to use this as an analogy to data centers, as you talk about large-scale investment.

We’re clearly the edge is coming, we’re not there yet, right? We’re not doing distributed edge compute everywhere. I think we’re years away from that, but I could see Africa not having the big weighty data center markets like we have in the FLAP or Northern Virginia, or Northern California. I could see that market supporting that edge distribution, any thoughts behind that, Andrew? And I know I’m kind of catching you with a conceptual question there.

Andrew Jay: I think you’re absolutely right. That there is … I don’t think there’s any way that they are going to be laying copper, or fiber in any big way, in any metropolitan areas. They’ll do the long haul with fiber obviously, but not locally. You’re absolutely right. Cell technology has leapfrogged and it’s really worked. I think that one of the interesting things I’ve been involved with recently is just the struggle they’re having with these remote locations on how to actually power the base stations, where they can’t get grid, or the grid so unreliable.

And so there’s some really interesting things happening over there, where they’re using some the new lithium ion batteries, combined with diesel generators when they have to, and then solar power. And it’s funny, I was talking to a company called, Power X who are looking at how you … And they’re actually testing it in Africa at the moment, how you can use artificial intelligence to get the absolute optimum blend of all of those different power sources to drive down costs and increase efficiency. So yeah, some interesting things happening that I think 5G will get there at some point. And I think, there mobile is the way to go, it’s the absolute test bed.

Raymond Hawkins: Well, Andrew, I want to be sensitive to your time, and I thank you for recording with us, and joining us, and we’d love to have you back when we can talk in more depth about other markets in Europe, and maybe even the Middle East, and a little bit of Africa. I think things changing in Africa also would be interesting. And we’ll do that for our next recording. Thank you so much for joining us and I look forward to talking to you again.

Andrew Jay: Thanks Raymond, it’s been an absolute pleasure. Good luck.

Raymond Hawkins: Take care, Andrew. Bye now.

Andrew Jay: Bye.