Celebrating 35 years of Telehouse: building a sustainable digital future

In this episode of “Making Connections”, host Nick Layzell reflects on 35 years of Telehouse with Mark Pestridge, Executive Vice President & General Manager of Telehouse Europe, and Sami Slim, CEO of Telehouse France. Together, they explore how Telehouse’s origins as a “carrier hotel” evolved into its position as a global leader in interconnected ecosystems.

The discussion highlights the growth of Telehouse’s data centre footprint, the industry’s shift to supporting AI and high-density workloads, and the ongoing importance of collaboration to tackle sustainability and power challenges. They also share how Telehouse’s approach to connectivity continues to shape the digital economy.

Listeners will hear how Telehouse’s focus on ecosystems, innovation, and sustainability has created a blueprint for the future of digital infrastructure.

Visit telehouse.net for more information

Transcript

Nick Layzell: 0:07

Hello and welcome to Making Connections, Telehouse's podcast, where we explore the IT challenges and the opportunities facing organizations today. I'm your host, Nick Clayzell, and today we're celebrating Telehouse's 35 years of evolution and expansion. I'm delighted to be joined by two key figures in Telehouse's growth story. Mark Pestridge, Executive Vice President and

Mark Pestridge: 0:26

GM Hi Nick. Yeah. So my name is Mark Pestridge. My role is to ensure that our customers have a seamless experience right across their journey with us. From the moment they start looking into our services through that whole kind of sales and contracting onboarding process, and then through their life cycle with us. So making sure that we are completely joined up and they have a fantastic experience from day one all the way through their life cycle.

Nick Layzell: 0:50

Thanks for joining me today, Mark. And Sami Slim, CEO of Telehouse Paris.

Sami Slim: 0:55

Hi Nick. Thanks for having me. I'm Sami. I am in charge of the French perimeter of Telehouse, basically mirroring Mark's role. But my job is to make our customers, our employee happy with the Telehouse operations in France.

Nick Layzell: 1:07

Over the past 35 years, Telehouse has grown from a single data center into a global leader in the data center business. And Sami, I think you've got a bit of a story about how it all began. It may be good to start with that.

Sami Slim: 1:18

Oh, yeah. The story of the birth of Telehouse is fantastic because it ties up the data center industry with the telco. The telco was what we call today the digital, telecom was the only thing people know about digital field in the eighties and seventies and eighties. And at some point when the encampment had been broken by the government, so the British telecom, the AT& T, the France telecom, the NTT came into the equation, the fact people coming into that field to compete with the incumbent had to benefit from a fair market condition and not to be completely overwhelmed by the big guys that have been subsidized by the government for four years. So one of the tricky question was, Where do we put our telco equipment? If I'm new, wanting to compete with British Telecom in the UK, where do I put my telco switches and routers to compete with these guys? Because if I go to their data center, which we're not called data center at the time, it was technical spaces, they will either, bill me a lot of money, or they will, protect their interconnection to go to customers, etc. So how do we solve that problem? And the genius idea came from two engineers from KDDI, so the parent company of Telehouse, saying, okay, we need a level playing field here. We need a place where people can compete freely and interconnect freely of the incumbent weight. And that should be a house for the telecom, so the Telehouse. And they came in with this idea, not in Japan, but in, in New York. In the U. S., they were in the wholesale division of KDDI, and they were into those problematic of how do people interconnect with the AT& T and stuff. And they said, okay, we will build a telecom house in New York, and we will replicate that model all around the globe to follow this trend of cutting the telco monopolies. And it was a massive success. And we are still living in out of that idea, original idea today, as we speak in London and Paris and many other cities.

Mark Pestridge: 3:18

Yeah, that's absolutely right. I think it's hard to believe that it was only 35 years ago that it all started. And as you say, it was really like an interconnection point for the main carriers to come together and exchange traffic back in, in the UK. If you think about what was Mercury in the day in BT, they were our first two customers in Telehouse North as it was. And they were there to exchange traffic. And it did take us a little while for. Other service providers to come on board and understand that they could interconnect in a neutral place. And it took us a long while to get going. And it was about 10 years between our first building, which was Telehouse North, which opened in 1990, about a year after the U S opening for us to actually build the next. That pace these days that's a really slow pace. But also you look at some of the challenge in those days, those challenges were very much about space rather than the power challenges we have today. So we were looking at less than half a kilowatt a rack. Can you believe that? Less than half a kilowatt a rack. And that was high density, right? So it was a real challenge around finding the space rather than the power. And that was really interesting. And it's also interesting to see that most of our competitors didn't really get going until the early 2000s. So that first 10 years, certainly in London was all about what we could do and try and get the market going and start to move. And I think about five years later, you guys opened up in Paris.

Sami Slim: 4:30

Exactly. So basically the playbook of Telehouse at the time was looking at the telco map and trying to find facilities where those fibers, converge in one point in each metro that we want to target and open a data center there. And it came out that when you do that, you go into the center of the cities. and you find office spaces that you need to retrofit as a data center. And again, it wasn't called data center at the time. We used to call it carrier hotel center. We're talking about office building in the center of the city. It was Metro in London. It was Junor in Voltaire. It was again, Chelsea in New York. So those office buildings have been retrofitted to become data centers. And they've actually lived a good life, I think, because their house has spent. 20 ish years in the very beginning of it, early stages, around those metro, because again, the aim of the business was to get interconnection in there. And this kind of constraint, being able to be as close as possible to the connection, put us into location that if we were to Replicate the business today. We would never think of opening a massive data center in the center of Paris. No one would do that. But at the time, that's what we did, how we did business. And I think it is still as of as we speak today, the strength of Telehouse. We have key assets that are feeding our new facilities around those key assets with connectivity. We're still getting the advantage of that rich history.

Mark Pestridge: 5:57

I think that evolution from what you call carrier hotel has really come onto what I would call now an interconnection hub, it used to be just service providers interconnecting, but now if you want to get to the cloud providers, if you want to get to SDNs, CDNs, ISPs, ASPs, they're all within Telehouse domain, be that in Paris or be that in London, be that in Canada, there's a huge amount of connectivity, but we've really changed now to being interconnection hubs rather than that, the old term, which was definitely carrier hotels. And maybe if we jump

Nick Layzell: 6:23

back then to, obviously 96, T. London was open for six years by that time in the market, and obviously the market was changing rapidly as well. The internet was, that, that ubiquitous thing that everybody has access to now is such a massive part of what we do. It was just really starting to take up, and the ideal location for companies that were internet suppliers or internet providers following the deregulation. The ideal place was these. Carrier hotels as we've termed.

Sami Slim: 6:49

So once you started the engine by making the carriers come into the same hotels, carrier hotel, the rest of the ecosystem that needed the carrier to go to the eyeball came into the same hotel. So the first trend was finance. Finance people wanted to be as close as possible to the telecoms because they needed, low latency. They need to say, so they say, okay let's not, reinvent the wheel. There, there are these. Things that are called carrier hotel, they're very good for us. Let's go there. So finance came in right after it was media and then the enterprises and then the cloud came in and then now the AI, et cetera. So you can see that the layers of the digital economy are building up and the same foundation, which is the carrier hotel for us, very simple reasons that as long as the physics say that nothing beats the speed of light, nothing will replace those fibers. Thank you very much. Because you cannot carry the information faster than the speed of light. So we will probably have to turn out our legacy data centers into huge aquariums in the future. But they will need to still be there because the fiber is in there and because nothing can go faster than light. That's what's happening. An industry catching up on the same facility with the same operators, such as Telehouse, building and keeping their strength, basing on their ability.

Nick Layzell: 8:09

And I'm sure that word ecosystem probably didn't exist back in the mid 90s. I don't remember hearing it back then, but it's a, it's almost a self fulfilling thing, isn't it? The more you build, maybe like a snowball effect or something, the more you build, the more facilities that you have, the more of those layers that come in, the more attractive it becomes for all types of businesses

Mark Pestridge: 8:27

to host there. Yeah, absolutely. I think it becomes a kind of symbiotic relationship if you like. So if you look at it from a content perspective, the content providers, they need to get their platforms to the eyeballs if you like. So they need the service providers to provide that connectivity to the residents, to the businesses, all those kinds of good things. But at the same time, the service providers need those content providers so they can provide those services to their customers as well. So it's really very much a real symbiotic relationship and both companies can be successful on the back of that. Because it's a really, it's a much simpler thing to do from a security point of view, an operational point of view, from a resilience point of view. It means that they can deliver what they need to deliver. But you're right. Ecosystems work was definitely not a word in 1990. But now it's the heart of what we do.

Nick Layzell: 9:07

Looking back, we're talking mid 90s and then coming up to the millennium, of course. So what are your thoughts on how technology then has driven through 2000 and beyond?

Sami Slim: 9:17

That's a very important question, I think. Probably we can talk also about it when we speak about the future, but so the data center has been the late part of the technology to evolve because basically the workload gets coded by, it people and then hardware companies get those workload to be. Run over some type of hardware, and that hardware needs to be somewhere, which is a data center. And the data center need to adapt to that hardware. So in the very beginning when we were a carrier hotel, we're basically hosting switches and routers. Those are easy machines. By easy machine, I mean they are switches. They send light right and left. Right and left. Okay. So from density point of view, from, cooling point of view from it, it is manageable to a certain extent. But once those layers that we were speaking about came in finance. media, etc. New hardware came in because the workload is no longer just switch the packet from right to left. It became, we need to compute in there, we need storage in there, we need some intelligence in there. So then some new hardware came in, which are the servers. And in the 90s, the server are things that we are not used to in a data center like ours. By the way, it's when the server came into our facility that we have ditched the word carrier hotel to become a data center. Because data came in, it's not only carriers, it's the data now. So here's Telehouse hosting servers now. And the 0. 5 kilowatt per rack now is becoming 4 kilowatt per rack, which was huge at the time. How are we going to manage? And at the time, the problem was not cooling, it was electricity. We used to have pdu, electrical pdu that are only handling one or two kilowatt. Now we need to go to four kilowatt. Wow. How are we gonna do that? So we invented the three phases, the one phase, et cetera. So this is how our. base of technology have evolved. So the workload evolves, which leads to a evolution in the hardware, which then lead to an evolution in the data center.

Mark Pestridge: 11:14

Yeah, no, completely agree. It's really interesting because there have been various different kind of evolutions as we've gone through our journey over the last 35 years. You talked about from interconnection with hubs and routers and switches to servers and compute. And I guess now. I guess we'll talk about it a little bit later, but even now we're talking about AI and we're talking about, base level 30 kilowatts, but we're talking about 100 kilowatts of racket. It's the same sort of problem because 0. 5 to 4 is the same as 12 to 100. We're dealing with the same problem. So we will fix those problems, but it's just it's really interesting how the industry continues

Nick Layzell: 11:46

to evolve. Sami touched on an interesting point there and, the technology or the, Business requirement maybe as it has led to the development of the data centers, but maybe a tricky question, I think, but but I'm going to pose it anyway. So from a data center perspective, looking back, where were we at that time building to prepare for what might happen in five years, or we didn't know what was going to happen in five years time?

Mark Pestridge: 12:08

That's a great question. I think we didn't have absolute certainty. I think that's for sure. I think like anything, you have to place certain bets. And I think we, we knew that. Demand from an electricity point of view or a power density point of view was going to go up. I guess we just didn't know which way and how fast it was going to go up. We could see the trajectory. We could see it absolutely increment. And it's all about talking to your customers as well. And they were all telling us that absolutely it's going to go up and up. And that is a real challenge for our business, which is if you need to build a data center today, it'll probably take us three years to go through planning before actually goes operational, you're talking about a different demand from customers in three years time than you are today. And that's a huge challenge that we have to think about. But in answer to your question, I think we didn't know exactly, but we were fairly confident that those densities were going to go up and up. I don't think if you'd asked us in 2000, if we ever thought we'd be dealing with 150 kilowatts a rack, we'd have probably laughed, but. But here we are.

Sami Slim: 13:02

Yeah there's one thing I would add to that, which is that the original intuition of Telehouse, which is building data centers around ecosystems and connectivity is still the anchoring fact in this evolution, meaning that whatever the hardware technology, the density, et cetera, is going, they are still asking for those ecosystems, for those cross connects. Your 100 kilowatt rack will still need to be close to the ecosystem and to the fibers. Meaning that we have figured out something very early in the process, which is that connectivity drives everything. Ecosystem drives everything. Interconnection drives everything. The rest is a byproduct that we need to adapt to. But it's not the other way around. It's not reinvent the 100 kilowatt and then let's figure out where we put it. No. We Build ecosystem, and then we make those ecosystem evolve to the technology.

Nick Layzell: 13:51

That original idea still holds true. That first step on the road to establishing these carrier hotels has morphed, has changed, has evolved, as we said earlier, but still completely holds true. And not every data center is the same. Maybe explain a little bit about the difference between that Telehouse model and another data centers that might be serving more of the kind of pure

Mark Pestridge: 14:12

compute. Yeah, although I think that. I'm absolutely confident that interconnection is going to be fundamental to our industry for the next 35 years plus. I have no doubt about that. I think we have a fundamental role to play in making sure that people can. Create these ecosystems. They can run their business securely, effectively, low latency, all those kinds of good things. So I'm really confident, but I do think you're right. There is a, another market, if you like, which for one of a better expression can be out of town where you're not as reliant on, or you don't need as much connectivity, but you do need a lot of space and more importantly, you need access to power. It's very difficult to build huge warehouses in the center of London, Paris, Toronto, Frankfurt, but you, if you can take those out of town and there is. Grid power available. I think you can start building those big warehouses where you can do all the compute stuff, but you will need a connection back into an interconnection hub or a carrier hotel or whatever you want to call it to facilitate all those connections. But I think that world is evolving pretty fast and it's all about access to power and it's all about access to. to space. But more importantly, I think it's about making sure that you can create them in a sustainable way as well. And I think that's another thing that's really driven the changes in our industry over the last five to 10 years as well, making sure that we can build and manage responsibly.

Sami Slim: 15:25

This is an important point for the listeners to understand what's the difference between Telehouse and other data center provider. A lot of new. Data center developers are coming from the real estate world. That says something. We're coming from a technology company. We're still a technology company. We're not doing real estate, right? Because we're still about switching packets and, building ecosystem, even though we're adapting this new compute paradigm. But we're not into building warehouses for machines. Other people's business, and they are doing it very well, huge warehouses out of the town, but they will still need us because to make their data transit somewhere, or for the AI to make the inference after the training, they would need us for the inference, even though we are not specialized in the training. And this is something that is, I think, one of the singularity of Telehouse, that we are still faithful to our very original principles. Promise, which is building ecosystem, even though we are, evolving, but we're not a real estate company, we're still a technology company.

Nick Layzell: 16:20

Sure. I think one of the other things that I wanted to touch on actually, and maybe Mark Mark mentioned briefly sustainability, obviously as we move forward in time and as we get, in, in more recent times, things like sustainability have become key. But just before that, I wanted to touch on resilience. Many years ago when we started, I don't think resilience or redundancy or were probably top of the agenda when the finance companies started to make more use of co location and data centers, the requirement for resilience ramped up, and I think that probably further accelerated the ecosystem looking for diversity of of providers. Any thoughts on how we've seen resilience or redundancy develop over the last 15 years or so,

Sami Slim: 17:04

I think the best thing that happened over the journey of Telehouse is the arrival of our competitors. Because, again, we came into this world over the promise of the government's breaking the monopolies on telcos, right? Which is the promise of giving the internet users the choice to choose. Of, who you want to work with and the resilience that comes with this choice because, you don't place your eggs in the same thing, right? You're building this kind of architecture on the IT. When our competitors come in, they basically propose to the market some resilience in each metro that we are operate with the Telehouse. So in the historical markets of Telehouse, London, Paris, or New York, you look at Telehouse Traffic weight, it's 50 percent of the Global peering traffic. And the sum of all the rest is 50%. So they are all basically building the resilience over the Telehouse. And we're doing that for our competitors and not another metros, meaning that it made the market bigger for everybody. And it gave the consumer the choice, which data center I want to use, but especially it gave them the tools and the industrial tools to replicate, be more resilient, more. And I think so now we're speaking about three availability zone in each metro. So it's not only two. Ps which was the norm 10 years ago. Now you need three, in, in a single metro to make sure that your critical workload is still running. So we can see that it has been beneficial to, to, to the entire industry, of course to data house as well. And we are still keeping, again, this promise of an open, interconnection. world and we're even accepting our competitors in our data centers because we want them to be part of our ecosystem. We want our customers to be able to connect to the IX of some of our competitors. Again the internet definition is a network of networks. So you cannot build walls or, boundaries on it. You just need to open it up.

Nick Layzell: 18:57

And now as we move towards the present time, there are other factors in running and operating data centers, giving customers that always on, always available, instantly available technology experience. But we need to consider sustainability. We need to consider how we move forward in the most efficient, least damaging way. Are we seeing customers bring sustainability more into focus?

Mark Pestridge: 19:21

Absolutely. I think quite rightly I think we, we have an obligation to the future of this planet to make sure that we do everything we possibly can to protect the future for our children's children, et cetera, et cetera, we absolutely have to do that. But I think from a business perspective it's an absolute. Essential from not only our customers, our employees as well, and our shareholders, they expect us to do everything in the most sustainable way that we possibly can. And I think there's a number of facets to that. I think, we've already talked about the fact that we've been in the industry for 35 years. So we have some inherent challenges that we have some buildings that are 35 plus years old. So they weren't built for the current challenges we have around density nor the current, Changes in climate. So we have to invest huge amount of capital to make sure that we're retrofitting those data centers, to make sure that we're in line with demands of not just our customers, but regulators and so on. I think that's a huge investment for us to do that. I think when we're building new data centers, we're very conscious of everything that we have to do to make sure they're as effective and as. Efficient as they possibly can be. And a lot of cost goes into designing those data centers to make sure they are built as sustainably as possible. And also we've just taken on a project where we have retrofitted an existing data center. So a building that was 35 years old, it would have done more damage to the environment to knock that building down and start again, even though that would have been easier for us. To actually, to manage the whole whole transition. So we've taken on board a project where we've actually retrofitted a data center whilst keeping the shell in place. We've managed to recycle, I think something like 96 or 97 percent of the equipment that was actually in that data center. So really it's front and center for, of everything we do. And it's not just about. the environment. It's also about the work you do in your community as well. So we have a responsibility to the community in which we live and which we operate. We're in London Borough of Tower Hamlets, which is one of the most challenged boroughs in the whole of London. And we're doing a lot of work with local charities to make sure that we're trying to give people in the borough as much opportunity as possible, whether that's through the donation of laptops, whether that's through actually providing Money through the equipment that we recycle or whether that's inviting them on open days to see what opportunities we could possibly give them within our facility. So it's huge. I think, I don't know if you agree, Sami, I think it's the most important thing we're doing at the moment to making sure that everything we do is as sustained as it was, we possibly can do. And just one last thing. I think that some of the things that we do, we talked about ecosystems. So if you remember when COVID hit in March, 2020, suddenly everybody had to work from home, but they had to collaborate. Where did they do that? They did that in data centers like Voltaire or Toronto or North. That's where that happened. And what did that mean? That mean there were no planes in the skies. People weren't driving cars. All facilitated by the fact that we had the ability to create collaborative and eco friendly ways of working.

Sami Slim: 22:06

Oh, spot on Mark, with success comes responsibility. So digital infrastructure and digital world has been very successful lately. And with that, we are under more scrutiny than we were in the past. When I started at Telehouse, people didn't even know what a data center mean. What is it? What are we talking about? What's this job you're having? Today, you say data center, ooh, you guys are consuming power and polluting and you guys are polluting. bad people, right? But this is a great challenge for us. And this is a good test. A company is the intersection of four shareholders, which are customers, of course, first, before everything, shareholders, detaining the capital, employees, and the community. The government, the regulators, and those four key stakeholders for any company are now asking for responsibility. All of them. We cannot hire someone today without explaining how we will make some effort for sustainability. Can you imagine that? This was not the case, Nick, just a few years ago. Today, when you interview someone, he just asks you, What are you guys doing, actually, to be sustainable in the future? Thank you. I'm not speaking about the customers who are asking for reporting. I'm speaking about shareholders who need to report their CO2 emission and communities. They are now challenging us over water. Why are you guys using water in your data centers? Water is for the living, not for the machines. They are right. We need to be responsible on that discussion. We need to be upfront about it. And as mentioned by Mark, we need to invest because we're talking about a lot of money, a lot of money to make those promises, life and making our industry sustainable. Otherwise the community, those four stakeholders, Either you won't have employee to work, or you won't have customers, or you won't have shareholders, giving you capital to invest, or you will have government shutting you down. So you need to be responsible over your CO2 emission, over your water usage, over your energy usage, and you need to be transparent. And this is exactly what we are doing right now, massively at Telehouse, all around the globe.

Nick Layzell: 24:11

I think it's such an important thing. Owning operating data centers is, of course, it's not just about the technology now. And I don't feel it's even a matter of balancing those different things. It's addressing all of them at the same time and with equal importance and rightly it's so important that we run and operate. In a sustainable way, because we're setting the tone now for what happens in five and 10 and 15 and 20 years time, at least, of course, we need to deliver the resilient, always on service for our customers, for their customers, the consumers. But doing that in line, managing all of those other concerns, and starting to lead the way in how we can address these challenges in the future is very important.

Mark Pestridge: 24:53

I completely agree. I think both of those things are minimal expectations of the four group of people that Sami just talked very eloquently about. So you have to operate with 100 percent uptime. But you have to do it in a sustainable way. There's not one trade off against the other. You have to do both of those as a minimum expectation. And I think that's the challenge that we have as a business, but that's where the investment comes in 24 by seven, a hundred percent uptime plus making sure you're doing everything in a sustainable way as you possibly can.

Nick Layzell: 25:19

Mark, you mentioned there are challenges. One of the challenges, and maybe the challenges as we look forward or now even, is around skill level, staffing, resource, the expectations of people that work in the data center industry or want to work in the data center industry. And it would be fair to say, Sami, that attracting the right talent, attracting the people that are passionate and exciting about data center currently is a challenge.

Sami Slim: 25:40

It is. So there's obviously in Europe, at least. A scarcity of talent around our jobs. And there is a challenge about the promise that we're giving these guys. We're seeing lots of talents, for example, leaving the carbon dense industries like petroleum or, oil, et cetera, because they think they are hurting the planet while working there, which I can understand. respect and relate to that. Now we're seeing those questions, even in our interviews. We asked the candidates, do you have questions for us? And the number one question that I'm getting is what are you guys doing for the planet? How are you making your business sustainable? And I find myself in a lot of those interviews, like convincing those young people about the massive industrial challenge that They can have if they join us by decarbonizing our industry, making it sustainable, et cetera, et cetera. And, defending our industry as well, because you need to be fair. The workload of our customers have increased by 500 percent over the 10 years. They're the consumption of our data centers have only increased by 6%. So we have been massively efficient, but it's not enough. It's not enough. We need to do more. And my speech to them is that if you're ambitious, If you want to save the planet, join us. You have a lot to do here to decarbonize the industry, to make the water usage to zero, to, et cetera, et cetera. And this is how we're trying to solve this equation. We're saying to talented, ambitious people, you want to have an impact, come join the digital infrastructure world because there is a lot to do there. And we're only in the beginning of the journey.

Nick Layzell: 27:12

And another issue may be the pace of the data center industry, huge developments and a huge amount of investment going into data centers to power this digital world. Our mobile phones our streaming services our banking services. So Mark, there's probably a lot of competition out there in

Mark Pestridge: 27:28

terms of the labor markets as well. I think there definitely is. I think we, we talked a little bit earlier on about Some of the new kind of out of town data centers that are getting a lot of investment, the data center world per se is viewed as a very interesting industry to invest in private equity funds, venture capitalists, et cetera. So we are seeing a lot of new entrance to the market, but there's still a finite pool of existing talent, if you like. So that's getting pulled around the industry. And I think it's also fair to say that a lot of the talent, particularly from an engineering perspective is becoming more and more mature. So we need to bring. People in that are younger. And I think what I love about the young people is they have new ideas. They don't just say, because you've done it this way for 35 years, you should carry on doing it. They challenge you. And to your point, Sami, about the questions they ask you in interviews, we do need to be challenged. I think what's got us to where we've got to over the last 35 years is not going to be good enough for the next 35 years. And we need People to really disrupt and challenge us. So I think bringing those, that new talent on board is a really important thing to do, and it's again, along with kind of sustainability and making sure we're doing all the things we need to hit our operational targets. Attracting, retaining, and developing. is front and center of what we do. I think there's a huge amount of work that, that we can do about publicizing about the importance of our industry. I know in the UK, since the Labour government came in, we've now been designated critical national infrastructure. That's hugely important. And the youngsters coming through can really help us develop that and evolve that and they can see how important we are. Not just to The social media world, but also to the, to businesses as well. And what an important function we play. So I think all that investment in apprenticeships in doing open days where you can really talk about not just the technical opportunities within a data center company, but some of the back office functions and how important it's going to be really important things to do. We in London have an excellent HR team. Who've got a really strong apprenticeship scheme that they're working on at the moment. And we've got some really good HR. Young graduate engineers joining our organization that has got to be the way forward, but it again inquires investment in class time and making sure that you're putting it front and center everything that you do. We

Nick Layzell: 29:31

recently held an open day actually in line with our charities who I wasn't there unfortunately I wish I was because the feedback I had was that there were some very excited. Young people that had come to the data center, the questions they were asking and the level of interest was phenomenal. So we need to harness that, don't we? And help them develop that interest in, in, in what we're

Mark Pestridge: 29:51

doing. No, absolutely right. And the reason that you and I weren't invited to that is we're way too old to be able to selling it to them. And we want to. We want people to see that it is a dynamic young industry. So having people like you and Munich around probably doesn't really help that. So that's the reason we weren't allowed to go there. Thanks. I'm putting myself in a podcast

Nick Layzell: 30:06

and Sami then, and Mark, I'm not going to ask you to make predictions because. We said earlier, I think it's very difficult to see what's coming five years down the line. But just generally, what do you see coming next in our industry, Sami? What do you think the future of both the data center industry and Telehouses?

Sami Slim: 30:24

We are going to hit a massive challenge with energy. We as an industry, not only Telehouse, okay? I think We underestimate how this is challenging because when you look at the AI model right now and how they are developed, Meta is doing a great job at opening data. You can read on their website, the evolution of their, you look at Lemma 1, Lemma 2, Lemma 3, Lemma 4 models. And from each one, you multiply by 100, even 1000 for the Lemma 5, the number of GPU, NVIDIA GPU they use to train just one single model. And they say, Actually, Lama 7 is coming, and the same factor is coming. So we do not see a limit to the number of GPU those things will consume, meaning that we do not see a limit to the amount of power that these models are going to require. I think this is a massive challenge for us as an industry, how to manage physically this process of transforming energy to digital usage without having an actual limit, coming to us. When I think about it, if we manage to solve it, we will make the evolution of our industry sustainable. So that's, as an industry, what we will have to manage in the next couple of years. But as Telehouse, we've made a very clear strategic choice. Strategy in business school, when they teach it to the young people we're talking about, it's about saying no. No, it's about choosing, right? And we said no to the real estate business. type of data center at Telehouse. We said, okay, we are about ecosystem. For example, for the AI, we're about inference. We're not into the, in the training of the model, which is she go at factories that we, we cannot do it. This is not our job. We are here to build Campuses like Docklands, hundreds of megawatts, but that's our max. So we made this clear choice that gave us a mission, which is continue providing our customers with space and power that is fit for usage that needs interconnection. But still, even in that way, we will have to, we will face some difficult decisions. For example, in Paris, We have decided to close an old facility, and that is a connectivity rich facility, Téléhouse 1 in Jeune Heure. It's its very first facility we've opened in Paris, very connectivity dense. But when we came into the conclusion that we cannot retrofit it properly for the sustainability challenge, and we cannot increase the power there, because again, the power hungry equipment are coming, we have told our customers very clearly, we have five years to migrate you guys to one of our other facility in Paris because we can no longer sustain that. That, that's a bold decision. Very few companies in the world are taking those kind of decision. And I'm not speaking about what Mark mentioned earlier, which is investing a lot of capital in existing fully let data centers just to refresh them. And make them, fit for the new age. So those are the two challenges I'm seeing. So managing the power, managing the legacy while continuing proposing to our customers fit for the ecosystem.

Mark Pestridge: 33:23

Yeah, I think that's absolutely spot on. I think we have to remain agile and flexible. I think there have been various different iterations or industry, as we talked about over the last 35 years from interconnection through the cloud boom. And now we're looking at the AI world. So we just need to make sure that we are being agile and flexible. There's two things I'd say. Firstly, that to your point, Sami, I think we cannot be all things to all people. We have got to be very focused on what we're good at. And secondly, we have to put the customers at the heart of everything we do. If we. understand what our customers need, why we need them, and we keep talking to them, listening to them rather than telling them, I think we'll be in a really good place moving forward because customers are absolutely front and center of what we need to do. So if we can understand what their challenges are, we can understand how and where our services can underpin them and help them be more successful, we'll be in a good place. But there will be many challenges ahead. And I think we're coming to probably the most challenging time in our industry for all the points that Sam has just made around power availability. Around sustainability it's a hugely challenging time, but if you have the right people, the right mindset, you put the customer front and center, you'll do well. And I'm confident that we as an organization will continue to thrive over the next 35 plus years. That feels like a

Nick Layzell: 34:32

big collaboration play, across our industry with our suppliers and partners, with our customers, like that collaboration is key so that we can all be heading towards a very bright future, a more sustainable

Mark Pestridge: 34:44

future. And I think there's a huge opportunity for us to collaborate with what we would call inverted commas, our competitors as well. I think a lot of us are facing the same challenges. I think we can collaborate together. There's a place for all of us to play. We offer different solutions. We offer different services. We offer different points of difference. So there's definitely an opportunity for us to collaborate with our so called competitors to, to make sure that we are, we're doing all the things that we need to and can help our customers.

Sami Slim: 35:09

This is a very important point because right now in the trajectory of development of Telehouse, we are leaving the tier one cities and going into the, what we call the tier two cities, like Marseille, like Thailand, Bangkok, etc. What we're seeing in those tier two cities now in 2024, which is completely different from what we, when we started in the eighties, is that the decision to build a Telehouse, a data center cluster. So for us and our competitors somewhere in shape, the entire city, it's energy arrivals, it's fiber distribution, the way we interconnect each cluster, etc. So meaning that when we build data centers in new cities now, we think of them as we were building airports or, huge railway stations. It's not like in the beginning of Telehouse, an old office building that we are retrofitting to a carrier hotel. This world is finished. This is now the world of massive infrastructure, very power hungry, that can shape an entire city around it. And I think this is where I think Mark is right by saying as an industry, we should collaborate as an industry. We should also discuss with governments and local officials say, Okay, your city now is a can be an important hub for the digital world. How do you want our infrastructure to fit into your society and your local community? That's a decision. We don't want to make is up. private company, because again, there is, there are too many stakeholders in that decision.

Nick Layzell: 36:36

It's certainly an exciting future. I think it's maybe more intimated that we're getting on slightly, but but what an exciting future for our industry, the young people. And thanks so much guys for having this chat today. I think hopefully the audience and and the listener found it very useful. Thanks very much, Sami. Thank you, Nick. And thanks, Mark. Thanks, Nick. Pleasure. Thanks, everyone, for listening to this episode of the Making Connections podcast, and I hope you'll join us for the next episode soon.