WEBVTT
00:00:00.160 --> 00:00:05.919
The data centre industry is moving faster than ever, and behind the scenes, some seriously smart people are making it all happen.
00:00:06.160 --> 00:00:12.960
Welcome to the Inside Data Center podcast, where we dive into the stories, insights, and innovations driving real change across the industry.
00:00:13.199 --> 00:00:19.839
From hyperscale campuses to game-changing technologies, we're talking to the people shaping the future of digital infrastructure.
00:00:20.000 --> 00:00:26.000
Whether you live and breathe data centers or just want to understand what's powering our connected world, you're in the right place.
00:00:26.239 --> 00:00:27.440
So we are back.
00:00:27.920 --> 00:00:36.159
Back at Data Cloud, back in Cannes, back in the sunshine, with only a slight delay to power this morning, which seemed like a pretty relevant conversation to be fair.
00:00:36.399 --> 00:00:39.359
Seeing how many umbrellas we can get to cover us, call us.
00:00:39.600 --> 00:00:40.240
Yeah, exactly.
00:00:40.320 --> 00:00:41.679
The theme of power and cooling.
00:00:41.840 --> 00:00:49.520
So we are here with Chris Pennington, Director Energy and Sustainability at Iron Mountain and Alfonso Salema, CEO of Excess Digital.
00:00:49.759 --> 00:00:51.039
Appreciate your time, guys.
00:00:51.200 --> 00:00:51.840
Massively so.
00:00:51.920 --> 00:00:52.560
So let's get into it.
00:00:52.799 --> 00:00:53.600
Happy to be here.
00:00:53.920 --> 00:00:54.479
Power.
00:00:55.039 --> 00:00:56.240
Power is the big trend.
00:00:56.320 --> 00:00:58.880
Power is the thing that everybody is talking about.
00:00:59.600 --> 00:01:00.960
Are we at capacity?
00:01:01.039 --> 00:01:04.239
Or does the data center industry need to build smarter?
00:01:05.040 --> 00:01:05.439
Cool.
00:01:05.840 --> 00:01:09.760
Well, I I, you know, I think access to power is the big challenge.
00:01:10.000 --> 00:01:23.599
Um, and I think part of finding access to power is understanding that the paradigm we've had in the past where we just show up, ask for this constant supply from the grid and expect to receive it is going to leave a lot of capacity stranded on the grid.
00:01:23.840 --> 00:01:34.879
So what we see now is that there's actually lots of capacity on many grids during much of the year, but unlocking access to that means we have to be able to find ways for our data center load to be flexible.
00:01:35.120 --> 00:01:43.599
So, how can we actually kind of change the shape of how we pull power from the grid to align with when capacity is available and when clean energy is available?
00:01:43.760 --> 00:01:57.200
And that's a lot of what we're focused on right now with some energy storage systems that can actually carry the load of the data center, the entire data center, for hours in order to unlock access to that stranded capacity that's there for the vast majority of the year.
00:01:57.519 --> 00:01:57.920
Nice.
00:01:58.159 --> 00:01:59.040
How funzai.
00:01:59.599 --> 00:02:04.959
Well, if if you recall, the first time we've spoken years ago.
00:02:05.200 --> 00:02:05.840
Many years ago.
00:02:06.000 --> 00:02:13.919
Many years ago, uh, which in this industry you corresponds like three decades, um we spoke about power.
00:02:14.080 --> 00:02:20.240
We spoke exactly about this power, uh, about the power issue, and uh it was uh out of the radar for most people.
00:02:20.479 --> 00:02:26.400
Um as you know, I come from a power background, so I've been looking into power grids for over 20 years.
00:02:26.560 --> 00:02:33.280
And the reality is uh that's a load of question because yes, we are capacity, at the same time, we're not.
00:02:40.319 --> 00:02:48.800
So when you build something, everything is built for peak peak load, peak consumption, which leaves most of things unused.
00:02:49.199 --> 00:02:59.919
You go to you go to a power line using EU codes, and um the TSO is going to ramp up that line for three times your actual need.
00:03:00.319 --> 00:03:03.039
You go to a substation, the same thing happens.
00:03:03.360 --> 00:03:16.319
So the TSO, of course, there's they're scared, they have no incentive to uh to optimize the grid access uh because all they have is punitive charges if there's a blackout.
00:03:16.560 --> 00:03:42.479
So um helping the TSO, I think it's the role of the data center industry also to help the TSO on how to make better choices, how to unlock more of the power, especially given how affordability concerns are uh affecting all of Europe and how uh data centers and that trend is starting in the US, their sense are perceived as pushing the affordability uh issue again, raising power prices to the ad consumers.
00:03:42.560 --> 00:03:47.919
So unless we help the TSOs with it, we're never going to use the full potential of the grid that we have now.
00:03:48.159 --> 00:03:51.120
Now, having said that, yes, there are constraints.
00:03:51.280 --> 00:03:59.520
There are effective constraints on the mark uh on the on the power grids uh in Europe that needs to be tackled, but those are almost surgical.
00:03:59.680 --> 00:04:05.360
It's not the last time I I saw there was a 300 billion plan to operate the European grid.
00:04:05.520 --> 00:04:07.039
That's wasted money.
00:04:07.439 --> 00:04:29.839
Because uh instead of moving uh loads to where that capacity is, we're trying to force feed in locations which uh are already strained and shouldn't have shouldn't have additional capacity because otherwise I mean after before a data center, the biggest equipment that humankind ever produced is the power grid.
00:04:30.079 --> 00:04:31.759
And it's a living organism.
00:04:32.079 --> 00:04:38.399
So we need to see it as such, and data centers need to adapt to to it in order to unlock, in order to grow.
00:04:38.639 --> 00:04:46.160
But it's a loaded question because we can spend time, we can spend the entire podcast just discussing one of those.
00:04:46.720 --> 00:04:55.360
You're absolutely right, it's a loaded question, and and and on that subject, let's get into kind of this idea of like innovation and power behind the meter.
00:04:55.680 --> 00:05:01.040
This is the the term everybody's talking about using behind the meter power to overcome greek constraints.
00:05:01.360 --> 00:05:02.560
Is this really a thing here?
00:05:03.279 --> 00:05:04.959
Oh, so I I live in the US.
00:05:05.040 --> 00:05:06.319
It's it's a massively effect.
00:05:06.639 --> 00:05:13.199
Um, of course, we have we have the luxury of uh access to plentiful, cheap natural gas.
00:05:13.360 --> 00:05:26.000
Um, and so the ability to use on-site power generation, using natural gas, of course, to just really not only accelerate the timeline to power, but gain certainty with when you'll get power.
00:05:26.160 --> 00:05:28.639
That's that's that's really where it comes in as being added value.
00:05:28.720 --> 00:05:31.519
And it's just we're seeing it at big scale, yeah.
00:05:31.600 --> 00:05:39.839
300, 400, 500 megawatts is not an uncommon size for an on-site generation plant that's being talked about or in design or in development.
00:05:40.079 --> 00:05:51.439
And it's it's it's a bit concerning, I think, for myself, just because it is very much a short-term solution that has potential long-term impacts.
00:05:51.519 --> 00:05:57.759
Um, not and not just environmentally, as you can imagine, for just you know burning natural gas, but just the financial commitments.
00:05:57.920 --> 00:06:03.040
So uh, you know, as a data center operator, we don't naturally go out and invest in power generation plants.
00:06:03.279 --> 00:06:13.519
And and yet these uh these contracts in order to bring this kind of capacity on site are typically 10, 15, maybe 20 years in length in order to get the average cost of power down to where it's reasonable.
00:06:13.759 --> 00:06:15.920
So um it's absolutely happening.
00:06:16.079 --> 00:06:20.560
It's it's happening for the sake of gaining certainty with when power will be available.
00:06:20.800 --> 00:06:28.879
It is uh hopefully a phenomenon that we see relatively short-lived, I think, because ultimately our customers, they all want to be connected to the grid.
00:06:28.959 --> 00:06:31.680
Like they want to be supplied by the grid for great reasons.
00:06:31.839 --> 00:06:35.600
I mean, the grid will over time become cleaner, like it is here in Europe.
00:06:36.079 --> 00:06:39.839
Um, it'll become more affordable because of clean power.
00:06:39.920 --> 00:06:46.240
And you just don't get those advantages when you build your power yourself behind the meter with natural gas on site.
00:06:46.480 --> 00:06:48.959
Um so yes, it's happening.
00:06:49.199 --> 00:06:51.680
Hopefully, it's a short-lived phenomenon.
00:06:51.759 --> 00:06:52.959
Um, but only time will tell.
00:06:53.040 --> 00:06:54.240
I mean he brushes down.
00:06:54.959 --> 00:07:00.879
We already had on a podcast, the first podcast we recorded about the customer and about how, yeah, what does the customer want?
00:07:00.959 --> 00:07:03.120
It's kind of the question we have to ask here, isn't it, really?
00:07:03.199 --> 00:07:08.079
Because it's okay having that solution, but if the customer doesn't want that solution, then it won't happen.
00:07:08.240 --> 00:07:09.439
Well, they may want it today.
00:07:09.600 --> 00:07:11.040
They may not want it five years from now.
00:07:11.279 --> 00:07:16.079
Yeah, but there's also you I mean, you talked on two very relevant points.
00:07:16.240 --> 00:07:26.079
Uh uh, the first one is yes, uh, so behind the meter is a means of a bridge to power, a bridge to grid uh process.
00:07:26.240 --> 00:07:36.560
So there's a big pipeline of projects that need to be operating as soon as possible, and you will need that behind the meter to speed it up because the grid timings are not the same as they just set the tiny.
00:07:36.800 --> 00:07:38.160
But that's one solution.
00:07:38.399 --> 00:07:46.560
The other one, uh, which is quite quite relevant and it's becoming insanely relevant, is the affordability.
00:07:46.879 --> 00:08:02.240
Affordability, uh the affordability of AI, the the sustainability of of the big companies, the big AI companies, which um, as you as you've probably seen from all the news that have been uh been coming, they're not they're not great.
00:08:02.319 --> 00:08:05.839
And they're not great because they're overpaying for power, they're overpaying for the location.
00:08:06.000 --> 00:08:09.439
So the TCO comes back to the to the forefront.
00:08:09.759 --> 00:08:13.040
So behind the meter can help on both.
00:08:13.360 --> 00:08:15.920
But you need to but you need a specific strategy.
00:08:16.079 --> 00:08:17.759
Am I using this as a bridge to grid?
00:08:17.839 --> 00:08:20.879
And then I use the grid and then I use that the same equipment.
00:08:20.959 --> 00:08:26.000
I use, I have uh I have a different use case for the long term, so that I reduce the TCO.
00:08:26.240 --> 00:08:33.440
Can I also use the grid as as a complementary once I have access to it, so I can further reduce the TCO?
00:08:34.480 --> 00:08:48.159
There's so much pressure now to deploy that the TCO question has stayed on the back seat, but now it's coming full trend at the front seat, and behind the meter uh will be one of the solutions to push it.
00:08:48.320 --> 00:08:51.679
Now, there's a difference between US and Europe, very big one.
00:08:51.919 --> 00:08:54.799
Uh, whereby in Europe you can have both.
00:08:54.879 --> 00:09:03.600
You can have grid and behind the meter co-living uh um again, that's that's uh an entire podcast where we can work on it.
00:09:03.759 --> 00:09:05.440
Um, but you can have both.
00:09:05.600 --> 00:09:06.960
In the US, you don't.
00:09:07.200 --> 00:09:12.559
Uh the regulator is still frowns uh on it, so it's hard to make that transition.
00:09:12.879 --> 00:09:25.519
Um, they're starting to have uh a little bit more understanding and see the value of it, especially with initiatives of bring your own uh power into the data center um that you're trying to build.
00:09:25.679 --> 00:09:31.120
So the the US is now doing going towards that direction, but it's not there yet.
00:09:31.200 --> 00:09:33.519
And while in Europe it it de facto is.
00:09:33.600 --> 00:09:45.600
And it probably the only way to make uh Europe competitive versus US is actually focused on bringing uh behind the meter power uh to lower the TCO.
00:09:46.080 --> 00:09:47.919
But make it more sustainable with that, huh?
00:09:48.159 --> 00:09:48.799
Yes.
00:09:49.200 --> 00:09:51.679
Because in Europe it's much cheaper.
00:09:51.919 --> 00:09:54.960
Uh renewables in in Europe are much cheaper.
00:09:55.120 --> 00:09:59.679
Uh so that's the de facto, the de facto enable enablement.
00:10:00.000 --> 00:10:02.480
Like you said, gas in the US is cheap.
00:10:02.720 --> 00:10:04.000
Gas in Europe is not.
00:10:04.240 --> 00:10:06.879
I was gonna say, like, do you think we'll see gas in Europe?
00:10:07.039 --> 00:10:10.399
Well, is it a viable solution in Europe?
00:10:11.279 --> 00:10:13.360
For some hedge hours, yes.
00:10:13.519 --> 00:10:16.720
And if you provide services to the grid, yes.
00:10:17.120 --> 00:10:23.519
Uh if you see it as your base load solution, then you're you're up to a big shock.
00:10:24.000 --> 00:10:27.440
Yeah, and yet also I agreed with I agree with you on that front.
00:10:27.600 --> 00:10:42.559
And also if you only utilize that power plant for some some peak capability, short time, like your your levelized cost of energy from that plant is massive because all the capacity pots, all the grid capa the the capacity on the gas system charges.
00:10:42.799 --> 00:10:55.360
Um so it's yeah, it's fundamentally um disadvantaged to try and solve the energy challenge or the power challenge individually, you know, behind the meter, one company each doing its own thing.
00:10:55.440 --> 00:10:57.440
It's a it's a sys it's a systems problem.
00:10:57.600 --> 00:11:11.759
So we need to think, you know, holistically, we need to think the grid, putting lots of power onto the grid, lots of resources, lots of different types, but fundamentally trying to you know, everybody kind of going doing their own thing individually, extremely inefficient.
00:11:12.080 --> 00:11:12.799
Completely agree.
00:11:13.039 --> 00:11:16.399
But when you look at other solutions, nuclear.
00:11:16.960 --> 00:11:23.840
We've seen France announce the 90 billion euros of investment this week, yesterday as we're in France, probably should talk about it.
00:11:23.919 --> 00:11:26.960
And a lot of that's related to the the access to nuclear power.
00:11:27.120 --> 00:11:30.799
So is again, is that I think there's there's you again it's another podcast.
00:11:31.120 --> 00:11:36.960
You can go down the community aspect of it as well and the perception of the of that side of power, like what do people actually think?
00:11:37.039 --> 00:11:37.840
Do they understand it?
00:11:37.919 --> 00:11:39.039
I think is is one question.
00:11:39.200 --> 00:11:42.240
But is Nucle is nuclear viable as a civilization?
00:11:44.399 --> 00:11:51.440
The la the last European big nuclear projects, they all all overshoot and are massively delayed.
00:11:51.759 --> 00:11:52.320
Let's talk about that.
00:11:52.480 --> 00:11:56.080
I think the UK will be building ours for yeah, decades and then should it be a big project.
00:11:56.720 --> 00:11:57.360
Exactly, exactly.
00:11:57.519 --> 00:12:02.000
So we go back to what we're discussing before the podcast, the the book, How Big Things Are Done.
00:12:02.159 --> 00:12:07.759
You have the bottom of the list, nuclear, it always overruns, it always uh gets delayed.
00:12:08.080 --> 00:12:18.639
So, I mean, there are a lot of people claiming that nuclear uh will see a new uh a new boo, a new boom, and there will be a solution.
00:12:18.879 --> 00:12:20.320
I'm not convinced.
00:12:20.480 --> 00:12:23.279
I haven't seen it, I haven't seen any evidence.
00:12:23.360 --> 00:12:23.919
Why is that so?
00:12:24.960 --> 00:12:26.559
I only have the opposite evidence.
00:12:26.720 --> 00:12:32.080
I've been doing this podcast for five years, nearly six years, and it's definitely been talked about the whole duration of that.
00:12:32.559 --> 00:12:34.480
And we and we haven't really progressed.
00:12:34.879 --> 00:12:43.759
How many times has the nuclear fleet in France has uh had to be intervened by government in order not to uh in order not to die?
00:12:44.399 --> 00:12:45.279
At least four.
00:12:45.360 --> 00:12:46.720
So I'm optimistic.
00:12:46.799 --> 00:12:53.919
I mean, I think in time it it will emerge as a viable solution, but I don't think it's gonna be in this behind the short term solution or something.
00:12:54.080 --> 00:13:09.200
You know, the small modular reactor concept, I think, is is a great avenue by which the technology can get commercialized and we can get a lot of advantages of of speed and scale and repetition in order to manufacture and bring down the cost and not be overrunning all the time.
00:13:09.360 --> 00:13:19.840
And yet ultimately, going back to what I just said, like individual companies trying to like solve this on their own with their own nuclear reactor, like it's possible.
00:13:19.919 --> 00:13:21.279
It's not the most efficient solution.
00:13:21.360 --> 00:13:24.399
So we need we need more nuclear on the grid, quite frankly.
00:13:24.480 --> 00:13:27.120
Um, and not a lot of kind of private wire nuclear.
00:13:27.679 --> 00:13:30.559
To your point earlier, everybody wants to be connected to the grid.
00:13:30.639 --> 00:13:30.799
Yeah.
00:13:30.960 --> 00:13:35.120
Some microgrids, nuclear power, SMRs at least, are transitional.
00:13:35.200 --> 00:13:35.679
Yeah.
00:13:36.000 --> 00:13:38.159
Is the transition worth worth the investment?
00:13:38.799 --> 00:13:39.440
Massive investment.
00:13:39.679 --> 00:13:40.799
Some massive investment, yeah.
00:13:41.120 --> 00:13:48.879
Just if you do, if you price it, if you price it just for the transition, how expensive running that data center would be.
00:13:49.039 --> 00:13:49.840
Insanely expensive.
00:13:49.919 --> 00:13:51.360
You need to pass that cost someone.
00:13:51.440 --> 00:13:51.600
Yeah.
00:13:51.759 --> 00:13:53.759
So you pass on to the final customer.
00:13:53.919 --> 00:14:05.440
So if people are already complaining that they're they're overshooting on token consumption and that their budgets on tokens are being evaporated on the first days of the month.
00:14:05.679 --> 00:14:10.320
Um if you multiply that cost by five, doesn't make sense.
00:14:12.399 --> 00:14:13.039
And that's the question.
00:14:13.120 --> 00:14:15.600
And that's the question that we need to ask.
00:14:15.840 --> 00:14:17.120
Nuclear can be a path.
00:14:17.279 --> 00:14:17.840
I agree with that.
00:14:18.159 --> 00:14:20.639
I've been industrial, I love the nuclear technology.
00:14:20.799 --> 00:14:27.039
I just haven't seen evidence of being cheap and meeting its deadlines.
00:14:27.200 --> 00:14:29.279
And that's what I I'd love to be proven wrong.
00:14:29.440 --> 00:14:29.600
Yeah.
00:14:29.919 --> 00:14:36.080
This probably brings us on to our next question perfectly as well, because we're talking about kind of power challenges.
00:14:36.399 --> 00:14:39.759
Are they the same in Europe as the US?
00:14:39.840 --> 00:14:42.080
Do we do it do we see the same things happening?
00:14:42.399 --> 00:14:44.399
I think you can even break that down by country as well.
00:14:44.559 --> 00:14:44.799
Yeah.
00:14:44.960 --> 00:14:47.360
Like the US very much.
00:14:48.159 --> 00:14:49.039
So it depends.
00:14:49.120 --> 00:14:51.440
There are you have some common themes at the top.
00:14:52.559 --> 00:15:00.720
Uh, but then uh as you start digging it up, even in the node, in a specific node, it's not the same problem as in another power node.
00:15:00.960 --> 00:15:07.039
So that idea that that you could have a one uh a one size fits all solution falls apart.
00:15:07.200 --> 00:15:09.039
Now, there are common themes.
00:15:09.279 --> 00:15:21.519
The affordability theme for the end customer, and if data centers coming into the grid will ramp up their cost bills, and uh it they will if they only use Powerbridge.
00:15:21.600 --> 00:15:27.120
So the bring your own power mentality is a great concept and it should be pushed for.
00:15:27.440 --> 00:15:37.440
Uh, but as you said, as a holistic system, as everyone connected to the grid and contributing to the same, not as an isolated, isolated means.
00:15:37.759 --> 00:15:40.000
Um, so that's the most common theme.
00:15:40.080 --> 00:15:48.879
But as you start coming down, like uh, I mean, Tolstoy used to say, all families aren't happy, but each family is unhappy on their own way.
00:15:49.039 --> 00:15:50.320
It's exactly the same thing.
00:15:50.559 --> 00:15:52.480
This is exactly the same thing.
00:15:52.799 --> 00:15:53.279
Interesting.
00:15:53.519 --> 00:15:59.279
Yeah, the we are electrifying so quickly that capacity is being consumed.
00:15:59.360 --> 00:16:03.039
And so I think I think that is one of those high-level themes we see everywhere.
00:16:03.120 --> 00:16:13.120
I don't I haven't I don't know of a single market where they're like, we're sitting around with like a plenty of capacity and like we're in a very fluid area where everybody wants to be, why is nobody doing it's like that doesn't exist.
00:16:13.440 --> 00:16:17.360
So um so I think yeah, capacity is a is a common issue.
00:16:17.519 --> 00:16:39.840
I think what's different is that we see because of the large-scale deployment of renewables across parts of Europe, like affordability is is proving to be very closely connected to um to flexibility, I would say, in terms of if you if you have the ability to use power at certain times a day, you can have very low prices.
00:16:40.000 --> 00:16:44.399
And so that's emerging as like an awareness, I think, more here in Europe than we see in the US.
00:16:44.559 --> 00:16:51.279
Um, just because a lot of our grids just simply don't have as much renewable penetration for the most part, is what we see across some of the markets in in Europe.
00:16:51.519 --> 00:16:54.320
And that's when you see like different use cases for different geographies.
00:16:54.799 --> 00:17:00.240
But yeah, we'll we'll build our learning, machine learning data center here, could be we'll build our cloud, could be.
00:17:00.879 --> 00:17:03.759
Or we just think about storage in ways that the load can be flexible.
00:17:04.240 --> 00:17:06.640
Which brings which brings back to your first question.
00:17:06.799 --> 00:17:08.799
Are we using the grid efficiently?
00:17:09.039 --> 00:17:12.799
And like like you said, uh, I mean, this phenomenon is not new.
00:17:12.880 --> 00:17:16.960
I mean, peak shaving has been used by many industries in the past.
00:17:17.039 --> 00:17:20.799
The data center industry was not using it because they didn't need to.
00:17:20.960 --> 00:17:21.119
Right.
00:17:21.279 --> 00:17:29.039
But now that we're forced to rethink of solutions that already exist and be more efficient about it, peak shaving is amazing.
00:17:29.279 --> 00:17:33.839
It can lower your your your power your power bill, but you need to be smart about it.
00:17:34.640 --> 00:17:40.480
And when you talk about community engagement, again, which will come up on every podcast we talk about here over the next few days, because it's such a big topic now.
00:17:40.559 --> 00:17:49.440
That's how we can help change the perception of the industry is it from the utilization of power to benefit the the community as well as the customer.
00:17:49.759 --> 00:18:03.119
And and that uh again, coming from power, uh, when I got into the data center industry, it was a shock on how little the data center industry does to to break the myths that are created uh along communities.
00:18:03.279 --> 00:18:07.440
I mean, the power industry has been doing that for decades uh successfully.
00:18:07.519 --> 00:18:16.000
Uh and we are, like you said, everyone is still trying to run their own thing instead of joining forces and let's start dismystifying.
00:18:16.160 --> 00:18:21.039
We're not here to increase your power bills, we're not here to drink all your water.
00:18:21.200 --> 00:18:22.480
We are more efficient than that.
00:18:22.640 --> 00:18:25.039
We know what we're doing, let's show that.
00:18:25.200 --> 00:18:31.359
And I think the data center industry needs to be much more vocal and united in doing this.
00:18:31.599 --> 00:18:34.720
And that's that's and that so that community engagement piece is more and more critical.
00:18:34.799 --> 00:18:36.240
More and more projects are being blocked.
00:18:36.319 --> 00:18:39.039
Uh especially now what it's hard in the US at the moment.
00:18:39.200 --> 00:18:39.359
Certainly.
00:18:39.519 --> 00:18:40.000
Certainly, for sure.
00:18:40.079 --> 00:18:46.799
I think I think I think one of the underlying challenges here is that um we are a business-to-business type industry.
00:18:46.960 --> 00:18:47.200
Yes.
00:18:47.359 --> 00:18:50.799
Um, and more than that, we're generally full of engineers.
00:18:50.960 --> 00:19:04.960
And and so like we're probably not we don't have the right DNA in most cases to show up and like listen to a community and talk about things that matter to them because we're, you know, we're gonna show up and talk about how evaporative cooling actually uses less water than a highly electrified data site.
00:19:05.200 --> 00:19:09.759
It's like, no, like people aren't going to understand our business the same way that we do.
00:19:09.920 --> 00:19:17.759
And so I think we just we have not done a great job and we're still trying to make progress to find ways to meet people where they are and have the right conversations.
00:19:17.920 --> 00:19:23.279
And I think that power is one of these places that we can bring real value to the community.
00:19:23.680 --> 00:19:26.400
We can show up and say, we can make your grid more reliable.
00:19:26.559 --> 00:19:32.400
We can actually put downward pressure on pricing for you by putting more power on that existing infrastructure.
00:19:32.559 --> 00:19:40.640
Like we can do these things that impact the kitchen table economics that I think will make an impact um to the community directly.
00:19:40.960 --> 00:20:04.960
And if you go to communities where um where large power uh stations used to lie, whatever type of power, uh, what you'll find is you'll find a willing community that understands economics of big infrastructure and they understand the impact, especially if that a power plant is about to be decommissioned, you have so many families which will will lose their jobs and the multiplying effect disappears.
00:20:05.200 --> 00:20:12.000
So and that's why I was saying the power industry did a good job uh setting up that zone.
00:20:12.240 --> 00:20:13.440
We should use that.
00:20:13.680 --> 00:20:18.160
We should use some of those examples to to relate better with the community.
00:20:18.480 --> 00:20:20.160
It's interesting that you say that.
00:20:20.240 --> 00:20:23.920
I was out at uh Terra Wolf in Buffalo like literally two weeks ago.
00:20:24.000 --> 00:20:29.440
I mean, just uh the scale of what they're building there is just absolutely phenomenal, like absolutely incredible.
00:20:29.680 --> 00:20:36.240
It's really interesting to hear how, first of all, they've taken this old disused coal plant and reconverted it into renewable power.
00:20:36.319 --> 00:20:45.359
And then, second of all, how the community surrounding that power station, like so many people have lost their jobs and they're they're at power distance, so they just went and rehired it.
00:20:45.519 --> 00:20:45.759
Of course.
00:20:45.920 --> 00:20:52.240
I and all of a sudden you hear this kind of you begin to realize that actually there is a really good catalyst with changes.
00:20:52.319 --> 00:20:57.359
And to your point, Chris, like I think um the industry's now increasing mindful of this.
00:20:57.440 --> 00:20:58.480
So I think that's one.
00:20:58.880 --> 00:21:02.160
We're a new industry, it's a really positive step forward.
00:21:02.240 --> 00:21:05.440
But it's certainly like Andy and myself have been talking about this for absolutely ages.
00:21:05.519 --> 00:21:07.119
Like, how do we bring people together?
00:21:07.200 --> 00:21:10.079
It was one of the reasons we founded the DCMC initiative.
00:21:10.400 --> 00:21:25.200
How do you actually bring marketers, public affairs, people, PR, comms people together to actually start addressing these problems of kind of a wider industry level instead of everyone running into the same brick walls in silos, because ultimately everyone's trying to achieve its own goals.
00:21:25.359 --> 00:21:29.920
I think it's about being responsible and being open enough to talk about it and say, we haven't got it figured out, right?
00:21:30.240 --> 00:21:33.519
And you'll find what do you find with communities, which is the most interesting?
00:21:33.680 --> 00:21:38.559
And again, like you said, we're mostly of us are engineers and we completely fail at that.
00:21:38.799 --> 00:21:45.039
But when you start running the exercise, the most interesting thing uh that you'll find is that communities, most of them, they just want to be heard.
00:21:45.200 --> 00:21:45.599
Yeah.
00:21:46.160 --> 00:21:49.359
Listen, literally listen to just want to be heard.
00:21:49.680 --> 00:21:51.039
You've got to listen to the community.
00:21:51.119 --> 00:21:53.519
I think that's a key part of this conversation.
00:21:53.599 --> 00:22:00.000
It's it's not about us telling communities what we're doing, it's about us listening to what their challenges are and solving their challenges.
00:22:00.480 --> 00:22:02.240
Again, that's what the industry is getting much better at.
00:22:02.319 --> 00:22:07.039
I think previously we just go into a community, tell them what we're gonna do, tell them how they're benefit, and they're like, hang on a minute.
00:22:07.839 --> 00:22:22.079
I I actually got told off by uh when I was doing one of these community sessions and say, we're sick of companies who come here, say that they'll do big things, they buy an ambulance, they take some photos, they leave away, and there's nothing here.
00:22:22.400 --> 00:22:25.759
Why don't we need the ambulance, we need other things.
00:22:25.920 --> 00:22:26.559
Yeah.
00:22:27.039 --> 00:22:28.240
It's a it's a good point.
00:22:28.400 --> 00:22:30.400
Less uh less talk, more action.
00:22:30.640 --> 00:22:38.160
A publicity to your point shouldn't be used as kind of a vehicle just to kind of say achieve the objective that you're trying to achieve from a commercial level.
00:22:38.559 --> 00:22:44.960
It actually needs to be responsible and leave something behind so you create, I suppose, a a legacy, the communities in which you're building it.
00:22:45.200 --> 00:22:49.039
And be prepared to have that be something very local, very specific to the whole community.
00:22:49.119 --> 00:22:51.200
There are many communities who might need that ambulance.
00:22:51.279 --> 00:22:53.519
Uh exactly, but but don't assume that everybody does.
00:22:54.480 --> 00:22:56.000
Or at least not just one anyway.
00:22:56.240 --> 00:22:56.400
Exactly.
00:22:56.640 --> 00:22:56.880
Exactly.
00:22:57.359 --> 00:22:59.440
So, guys, we are we are at time.
00:22:59.759 --> 00:23:01.839
Really, really appreciate you taking the time to speak with us.
00:23:02.160 --> 00:23:04.079
I have to ask you the marketing question.
00:23:04.319 --> 00:23:11.200
So uh so just in terms of uh the event, what's really excited you so far and what you're looking forward to?
00:23:11.519 --> 00:23:14.720
Well, it's just getting started, but the uh the award ceremony was last night.
00:23:14.799 --> 00:23:18.720
We won uh Best Global Beta Center Operator Year for the second year in a row.
00:23:18.880 --> 00:23:20.240
So that's been the most exciting thing for us.
00:23:20.559 --> 00:23:21.440
Retained your title.
00:23:21.599 --> 00:23:21.920
Well done.
00:23:22.000 --> 00:23:22.799
Yeah, nice.
00:23:22.960 --> 00:23:24.640
Yeah, multi-award winner now.
00:23:24.799 --> 00:23:25.440
Yeah, yeah.
00:23:25.599 --> 00:23:26.160
Uh yeah.
00:23:26.240 --> 00:23:27.599
So it's been it's been brilliant.
00:23:28.480 --> 00:23:48.480
So it's just started, but it's um I think the the thing that has struck me the most is to see uh first of all how this conference is growing, but also how some some some small companies, and we were talking about one uh a few minutes ago, that uh three, four years ago they were tiny, and now they're massive.
00:23:48.559 --> 00:23:50.240
They're throwing a party in a yard.
00:23:50.480 --> 00:23:58.799
I mean, the industry is growing a lot, and with it it cut you have new new companies coming in, being very successful.
00:23:58.960 --> 00:24:01.200
Uh, and it's great to see them here.
00:24:01.599 --> 00:24:01.839
Right.
00:24:02.000 --> 00:24:08.720
I think events are a great place to realize the scale of that growth because we're all in it, we live it, we breathe it all day, every day.
00:24:08.799 --> 00:24:14.160
You don't really notice it too much, and you come back a year later and you realize there's double the amount of the people, yeah.
00:24:14.240 --> 00:24:17.440
There's another flaw in the in the building, which I found out 10 minutes ago.
00:24:17.519 --> 00:24:22.480
Um, yeah, I think the scale of this industry is a lot off the scale now, hang it.
00:24:22.559 --> 00:24:25.359
And and they're they're already forecasting bigger next year.
00:24:25.599 --> 00:24:26.480
Are they really?
00:24:27.680 --> 00:24:28.720
But enjoy the event.
00:24:28.799 --> 00:24:29.839
Thanks both for coming on.
00:24:30.079 --> 00:24:30.240
Thank you.
00:24:30.400 --> 00:24:34.000
Thanks for the that's it for this episode of the Inside Data Centre Podcast.
00:24:34.079 --> 00:24:35.039
Thanks for tuning in.
00:24:35.200 --> 00:24:43.119
This show is brought to you by Data X and Data Exec, the recruitment partners powering global talent across the data centre and digital infrastructure world.
00:24:43.359 --> 00:24:47.200
If you enjoyed today's conversations, don't forget to follow, like, and subscribe.
00:24:47.359 --> 00:24:49.839
It really helps us grow the podcast community.
00:24:50.079 --> 00:24:55.440
Stay tuned for more stories, insights, and conversations with the people shaping the future of our industry.
00:00:00.160 --> 00:00:05.919
The data centre industry is moving faster than ever, and behind the scenes, some seriously smart people are making it all happen.
00:00:06.160 --> 00:00:12.960
Welcome to the Inside Data Center podcast, where we dive into the stories, insights, and innovations driving real change across the industry.
00:00:13.199 --> 00:00:19.839
From hyperscale campuses to game-changing technologies, we're talking to the people shaping the future of digital infrastructure.
00:00:20.000 --> 00:00:26.000
Whether you live and breathe data centers or just want to understand what's powering our connected world, you're in the right place.
00:00:26.239 --> 00:00:27.440
So we are back.
00:00:27.920 --> 00:00:36.159
Back at Data Cloud, back in Cannes, back in the sunshine, with only a slight delay to power this morning, which seemed like a pretty relevant conversation to be fair.
00:00:36.399 --> 00:00:39.359
Seeing how many umbrellas we can get to cover us, call us.
00:00:39.600 --> 00:00:40.240
Yeah, exactly.
00:00:40.320 --> 00:00:41.679
The theme of power and cooling.
00:00:41.840 --> 00:00:49.520
So we are here with Chris Pennington, Director Energy and Sustainability at Iron Mountain and Alfonso Salema, CEO of Excess Digital.
00:00:49.759 --> 00:00:51.039
Appreciate your time, guys.
00:00:51.200 --> 00:00:51.840
Massively so.
00:00:51.920 --> 00:00:52.560
So let's get into it.
00:00:52.799 --> 00:00:53.600
Happy to be here.
00:00:53.920 --> 00:00:54.479
Power.
00:00:55.039 --> 00:00:56.240
Power is the big trend.
00:00:56.320 --> 00:00:58.880
Power is the thing that everybody is talking about.
00:00:59.600 --> 00:01:00.960
Are we at capacity?
00:01:01.039 --> 00:01:04.239
Or does the data center industry need to build smarter?
00:01:05.040 --> 00:01:05.439
Cool.
00:01:05.840 --> 00:01:09.760
Well, I I, you know, I think access to power is the big challenge.
00:01:10.000 --> 00:01:23.599
Um, and I think part of finding access to power is understanding that the paradigm we've had in the past where we just show up, ask for this constant supply from the grid and expect to receive it is going to leave a lot of capacity stranded on the grid.
00:01:23.840 --> 00:01:34.879
So what we see now is that there's actually lots of capacity on many grids during much of the year, but unlocking access to that means we have to be able to find ways for our data center load to be flexible.
00:01:35.120 --> 00:01:43.599
So, how can we actually kind of change the shape of how we pull power from the grid to align with when capacity is available and when clean energy is available?
00:01:43.760 --> 00:01:57.200
And that's a lot of what we're focused on right now with some energy storage systems that can actually carry the load of the data center, the entire data center, for hours in order to unlock access to that stranded capacity that's there for the vast majority of the year.
00:01:57.519 --> 00:01:57.920
Nice.
00:01:58.159 --> 00:01:59.040
How funzai.
00:01:59.599 --> 00:02:04.959
Well, if if you recall, the first time we've spoken years ago.
00:02:05.200 --> 00:02:05.840
Many years ago.
00:02:06.000 --> 00:02:13.919
Many years ago, uh, which in this industry you corresponds like three decades, um we spoke about power.
00:02:14.080 --> 00:02:20.240
We spoke exactly about this power, uh, about the power issue, and uh it was uh out of the radar for most people.
00:02:20.479 --> 00:02:26.400
Um as you know, I come from a power background, so I've been looking into power grids for over 20 years.
00:02:26.560 --> 00:02:33.280
And the reality is uh that's a load of question because yes, we are capacity, at the same time, we're not.
00:02:40.319 --> 00:02:48.800
So when you build something, everything is built for peak peak load, peak consumption, which leaves most of things unused.
00:02:49.199 --> 00:02:59.919
You go to you go to a power line using EU codes, and um the TSO is going to ramp up that line for three times your actual need.
00:03:00.319 --> 00:03:03.039
You go to a substation, the same thing happens.
00:03:03.360 --> 00:03:16.319
So the TSO, of course, there's they're scared, they have no incentive to uh to optimize the grid access uh because all they have is punitive charges if there's a blackout.
00:03:16.560 --> 00:03:42.479
So um helping the TSO, I think it's the role of the data center industry also to help the TSO on how to make better choices, how to unlock more of the power, especially given how affordability concerns are uh affecting all of Europe and how uh data centers and that trend is starting in the US, their sense are perceived as pushing the affordability uh issue again, raising power prices to the ad consumers.
00:03:42.560 --> 00:03:47.919
So unless we help the TSOs with it, we're never going to use the full potential of the grid that we have now.
00:03:48.159 --> 00:03:51.120
Now, having said that, yes, there are constraints.
00:03:51.280 --> 00:03:59.520
There are effective constraints on the mark uh on the on the power grids uh in Europe that needs to be tackled, but those are almost surgical.
00:03:59.680 --> 00:04:05.360
It's not the last time I I saw there was a 300 billion plan to operate the European grid.
00:04:05.520 --> 00:04:07.039
That's wasted money.
00:04:07.439 --> 00:04:29.839
Because uh instead of moving uh loads to where that capacity is, we're trying to force feed in locations which uh are already strained and shouldn't have shouldn't have additional capacity because otherwise I mean after before a data center, the biggest equipment that humankind ever produced is the power grid.
00:04:30.079 --> 00:04:31.759
And it's a living organism.
00:04:32.079 --> 00:04:38.399
So we need to see it as such, and data centers need to adapt to to it in order to unlock, in order to grow.
00:04:38.639 --> 00:04:46.160
But it's a loaded question because we can spend time, we can spend the entire podcast just discussing one of those.
00:04:46.720 --> 00:04:55.360
You're absolutely right, it's a loaded question, and and and on that subject, let's get into kind of this idea of like innovation and power behind the meter.
00:04:55.680 --> 00:05:01.040
This is the the term everybody's talking about using behind the meter power to overcome greek constraints.
00:05:01.360 --> 00:05:02.560
Is this really a thing here?
00:05:03.279 --> 00:05:04.959
Oh, so I I live in the US.
00:05:05.040 --> 00:05:06.319
It's it's a massively effect.
00:05:06.639 --> 00:05:13.199
Um, of course, we have we have the luxury of uh access to plentiful, cheap natural gas.
00:05:13.360 --> 00:05:26.000
Um, and so the ability to use on-site power generation, using natural gas, of course, to just really not only accelerate the timeline to power, but gain certainty with when you'll get power.
00:05:26.160 --> 00:05:28.639
That's that's that's really where it comes in as being added value.
00:05:28.720 --> 00:05:31.519
And it's just we're seeing it at big scale, yeah.
00:05:31.600 --> 00:05:39.839
300, 400, 500 megawatts is not an uncommon size for an on-site generation plant that's being talked about or in design or in development.
00:05:40.079 --> 00:05:51.439
And it's it's it's a bit concerning, I think, for myself, just because it is very much a short-term solution that has potential long-term impacts.
00:05:51.519 --> 00:05:57.759
Um, not and not just environmentally, as you can imagine, for just you know burning natural gas, but just the financial commitments.
00:05:57.920 --> 00:06:03.040
So uh, you know, as a data center operator, we don't naturally go out and invest in power generation plants.
00:06:03.279 --> 00:06:13.519
And and yet these uh these contracts in order to bring this kind of capacity on site are typically 10, 15, maybe 20 years in length in order to get the average cost of power down to where it's reasonable.
00:06:13.759 --> 00:06:15.920
So um it's absolutely happening.
00:06:16.079 --> 00:06:20.560
It's it's happening for the sake of gaining certainty with when power will be available.
00:06:20.800 --> 00:06:28.879
It is uh hopefully a phenomenon that we see relatively short-lived, I think, because ultimately our customers, they all want to be connected to the grid.
00:06:28.959 --> 00:06:31.680
Like they want to be supplied by the grid for great reasons.
00:06:31.839 --> 00:06:35.600
I mean, the grid will over time become cleaner, like it is here in Europe.
00:06:36.079 --> 00:06:39.839
Um, it'll become more affordable because of clean power.
00:06:39.920 --> 00:06:46.240
And you just don't get those advantages when you build your power yourself behind the meter with natural gas on site.
00:06:46.480 --> 00:06:48.959
Um so yes, it's happening.
00:06:49.199 --> 00:06:51.680
Hopefully, it's a short-lived phenomenon.
00:06:51.759 --> 00:06:52.959
Um, but only time will tell.
00:06:53.040 --> 00:06:54.240
I mean he brushes down.
00:06:54.959 --> 00:07:00.879
We already had on a podcast, the first podcast we recorded about the customer and about how, yeah, what does the customer want?
00:07:00.959 --> 00:07:03.120
It's kind of the question we have to ask here, isn't it, really?
00:07:03.199 --> 00:07:08.079
Because it's okay having that solution, but if the customer doesn't want that solution, then it won't happen.
00:07:08.240 --> 00:07:09.439
Well, they may want it today.
00:07:09.600 --> 00:07:11.040
They may not want it five years from now.
00:07:11.279 --> 00:07:16.079
Yeah, but there's also you I mean, you talked on two very relevant points.
00:07:16.240 --> 00:07:26.079
Uh uh, the first one is yes, uh, so behind the meter is a means of a bridge to power, a bridge to grid uh process.
00:07:26.240 --> 00:07:36.560
So there's a big pipeline of projects that need to be operating as soon as possible, and you will need that behind the meter to speed it up because the grid timings are not the same as they just set the tiny.
00:07:36.800 --> 00:07:38.160
But that's one solution.
00:07:38.399 --> 00:07:46.560
The other one, uh, which is quite quite relevant and it's becoming insanely relevant, is the affordability.
00:07:46.879 --> 00:08:02.240
Affordability, uh the affordability of AI, the the sustainability of of the big companies, the big AI companies, which um, as you as you've probably seen from all the news that have been uh been coming, they're not they're not great.
00:08:02.319 --> 00:08:05.839
And they're not great because they're overpaying for power, they're overpaying for the location.
00:08:06.000 --> 00:08:09.439
So the TCO comes back to the to the forefront.
00:08:09.759 --> 00:08:13.040
So behind the meter can help on both.
00:08:13.360 --> 00:08:15.920
But you need to but you need a specific strategy.
00:08:16.079 --> 00:08:17.759
Am I using this as a bridge to grid?
00:08:17.839 --> 00:08:20.879
And then I use the grid and then I use that the same equipment.
00:08:20.959 --> 00:08:26.000
I use, I have uh I have a different use case for the long term, so that I reduce the TCO.
00:08:26.240 --> 00:08:33.440
Can I also use the grid as as a complementary once I have access to it, so I can further reduce the TCO?
00:08:34.480 --> 00:08:48.159
There's so much pressure now to deploy that the TCO question has stayed on the back seat, but now it's coming full trend at the front seat, and behind the meter uh will be one of the solutions to push it.
00:08:48.320 --> 00:08:51.679
Now, there's a difference between US and Europe, very big one.
00:08:51.919 --> 00:08:54.799
Uh, whereby in Europe you can have both.
00:08:54.879 --> 00:09:03.600
You can have grid and behind the meter co-living uh um again, that's that's uh an entire podcast where we can work on it.
00:09:03.759 --> 00:09:05.440
Um, but you can have both.
00:09:05.600 --> 00:09:06.960
In the US, you don't.
00:09:07.200 --> 00:09:12.559
Uh the regulator is still frowns uh on it, so it's hard to make that transition.
00:09:12.879 --> 00:09:25.519
Um, they're starting to have uh a little bit more understanding and see the value of it, especially with initiatives of bring your own uh power into the data center um that you're trying to build.
00:09:25.679 --> 00:09:31.120
So the the US is now doing going towards that direction, but it's not there yet.
00:09:31.200 --> 00:09:33.519
And while in Europe it it de facto is.
00:09:33.600 --> 00:09:45.600
And it probably the only way to make uh Europe competitive versus US is actually focused on bringing uh behind the meter power uh to lower the TCO.
00:09:46.080 --> 00:09:47.919
But make it more sustainable with that, huh?
00:09:48.159 --> 00:09:48.799
Yes.
00:09:49.200 --> 00:09:51.679
Because in Europe it's much cheaper.
00:09:51.919 --> 00:09:54.960
Uh renewables in in Europe are much cheaper.
00:09:55.120 --> 00:09:59.679
Uh so that's the de facto, the de facto enable enablement.
00:10:00.000 --> 00:10:02.480
Like you said, gas in the US is cheap.
00:10:02.720 --> 00:10:04.000
Gas in Europe is not.
00:10:04.240 --> 00:10:06.879
I was gonna say, like, do you think we'll see gas in Europe?
00:10:07.039 --> 00:10:10.399
Well, is it a viable solution in Europe?
00:10:11.279 --> 00:10:13.360
For some hedge hours, yes.
00:10:13.519 --> 00:10:16.720
And if you provide services to the grid, yes.
00:10:17.120 --> 00:10:23.519
Uh if you see it as your base load solution, then you're you're up to a big shock.
00:10:24.000 --> 00:10:27.440
Yeah, and yet also I agreed with I agree with you on that front.
00:10:27.600 --> 00:10:42.559
And also if you only utilize that power plant for some some peak capability, short time, like your your levelized cost of energy from that plant is massive because all the capacity pots, all the grid capa the the capacity on the gas system charges.
00:10:42.799 --> 00:10:55.360
Um so it's yeah, it's fundamentally um disadvantaged to try and solve the energy challenge or the power challenge individually, you know, behind the meter, one company each doing its own thing.
00:10:55.440 --> 00:10:57.440
It's a it's a sys it's a systems problem.
00:10:57.600 --> 00:11:11.759
So we need to think, you know, holistically, we need to think the grid, putting lots of power onto the grid, lots of resources, lots of different types, but fundamentally trying to you know, everybody kind of going doing their own thing individually, extremely inefficient.
00:11:12.080 --> 00:11:12.799
Completely agree.
00:11:13.039 --> 00:11:16.399
But when you look at other solutions, nuclear.
00:11:16.960 --> 00:11:23.840
We've seen France announce the 90 billion euros of investment this week, yesterday as we're in France, probably should talk about it.
00:11:23.919 --> 00:11:26.960
And a lot of that's related to the the access to nuclear power.
00:11:27.120 --> 00:11:30.799
So is again, is that I think there's there's you again it's another podcast.
00:11:31.120 --> 00:11:36.960
You can go down the community aspect of it as well and the perception of the of that side of power, like what do people actually think?
00:11:37.039 --> 00:11:37.840
Do they understand it?
00:11:37.919 --> 00:11:39.039
I think is is one question.
00:11:39.200 --> 00:11:42.240
But is Nucle is nuclear viable as a civilization?
00:11:44.399 --> 00:11:51.440
The la the last European big nuclear projects, they all all overshoot and are massively delayed.
00:11:51.759 --> 00:11:52.320
Let's talk about that.
00:11:52.480 --> 00:11:56.080
I think the UK will be building ours for yeah, decades and then should it be a big project.
00:11:56.720 --> 00:11:57.360
Exactly, exactly.
00:11:57.519 --> 00:12:02.000
So we go back to what we're discussing before the podcast, the the book, How Big Things Are Done.
00:12:02.159 --> 00:12:07.759
You have the bottom of the list, nuclear, it always overruns, it always uh gets delayed.
00:12:08.080 --> 00:12:18.639
So, I mean, there are a lot of people claiming that nuclear uh will see a new uh a new boo, a new boom, and there will be a solution.
00:12:18.879 --> 00:12:20.320
I'm not convinced.
00:12:20.480 --> 00:12:23.279
I haven't seen it, I haven't seen any evidence.
00:12:23.360 --> 00:12:23.919
Why is that so?
00:12:24.960 --> 00:12:26.559
I only have the opposite evidence.
00:12:26.720 --> 00:12:32.080
I've been doing this podcast for five years, nearly six years, and it's definitely been talked about the whole duration of that.
00:12:32.559 --> 00:12:34.480
And we and we haven't really progressed.
00:12:34.879 --> 00:12:43.759
How many times has the nuclear fleet in France has uh had to be intervened by government in order not to uh in order not to die?
00:12:44.399 --> 00:12:45.279
At least four.
00:12:45.360 --> 00:12:46.720
So I'm optimistic.
00:12:46.799 --> 00:12:53.919
I mean, I think in time it it will emerge as a viable solution, but I don't think it's gonna be in this behind the short term solution or something.
00:12:54.080 --> 00:13:09.200
You know, the small modular reactor concept, I think, is is a great avenue by which the technology can get commercialized and we can get a lot of advantages of of speed and scale and repetition in order to manufacture and bring down the cost and not be overrunning all the time.
00:13:09.360 --> 00:13:19.840
And yet ultimately, going back to what I just said, like individual companies trying to like solve this on their own with their own nuclear reactor, like it's possible.
00:13:19.919 --> 00:13:21.279
It's not the most efficient solution.
00:13:21.360 --> 00:13:24.399
So we need we need more nuclear on the grid, quite frankly.
00:13:24.480 --> 00:13:27.120
Um, and not a lot of kind of private wire nuclear.
00:13:27.679 --> 00:13:30.559
To your point earlier, everybody wants to be connected to the grid.
00:13:30.639 --> 00:13:30.799
Yeah.
00:13:30.960 --> 00:13:35.120
Some microgrids, nuclear power, SMRs at least, are transitional.
00:13:35.200 --> 00:13:35.679
Yeah.
00:13:36.000 --> 00:13:38.159
Is the transition worth worth the investment?
00:13:38.799 --> 00:13:39.440
Massive investment.
00:13:39.679 --> 00:13:40.799
Some massive investment, yeah.
00:13:41.120 --> 00:13:48.879
Just if you do, if you price it, if you price it just for the transition, how expensive running that data center would be.
00:13:49.039 --> 00:13:49.840
Insanely expensive.
00:13:49.919 --> 00:13:51.360
You need to pass that cost someone.
00:13:51.440 --> 00:13:51.600
Yeah.
00:13:51.759 --> 00:13:53.759
So you pass on to the final customer.
00:13:53.919 --> 00:14:05.440
So if people are already complaining that they're they're overshooting on token consumption and that their budgets on tokens are being evaporated on the first days of the month.
00:14:05.679 --> 00:14:10.320
Um if you multiply that cost by five, doesn't make sense.
00:14:12.399 --> 00:14:13.039
And that's the question.
00:14:13.120 --> 00:14:15.600
And that's the question that we need to ask.
00:14:15.840 --> 00:14:17.120
Nuclear can be a path.
00:14:17.279 --> 00:14:17.840
I agree with that.
00:14:18.159 --> 00:14:20.639
I've been industrial, I love the nuclear technology.
00:14:20.799 --> 00:14:27.039
I just haven't seen evidence of being cheap and meeting its deadlines.
00:14:27.200 --> 00:14:29.279
And that's what I I'd love to be proven wrong.
00:14:29.440 --> 00:14:29.600
Yeah.
00:14:29.919 --> 00:14:36.080
This probably brings us on to our next question perfectly as well, because we're talking about kind of power challenges.
00:14:36.399 --> 00:14:39.759
Are they the same in Europe as the US?
00:14:39.840 --> 00:14:42.080
Do we do it do we see the same things happening?
00:14:42.399 --> 00:14:44.399
I think you can even break that down by country as well.
00:14:44.559 --> 00:14:44.799
Yeah.
00:14:44.960 --> 00:14:47.360
Like the US very much.
00:14:48.159 --> 00:14:49.039
So it depends.
00:14:49.120 --> 00:14:51.440
There are you have some common themes at the top.
00:14:52.559 --> 00:15:00.720
Uh, but then uh as you start digging it up, even in the node, in a specific node, it's not the same problem as in another power node.
00:15:00.960 --> 00:15:07.039
So that idea that that you could have a one uh a one size fits all solution falls apart.
00:15:07.200 --> 00:15:09.039
Now, there are common themes.
00:15:09.279 --> 00:15:21.519
The affordability theme for the end customer, and if data centers coming into the grid will ramp up their cost bills, and uh it they will if they only use Powerbridge.
00:15:21.600 --> 00:15:27.120
So the bring your own power mentality is a great concept and it should be pushed for.
00:15:27.440 --> 00:15:37.440
Uh, but as you said, as a holistic system, as everyone connected to the grid and contributing to the same, not as an isolated, isolated means.
00:15:37.759 --> 00:15:40.000
Um, so that's the most common theme.
00:15:40.080 --> 00:15:48.879
But as you start coming down, like uh, I mean, Tolstoy used to say, all families aren't happy, but each family is unhappy on their own way.
00:15:49.039 --> 00:15:50.320
It's exactly the same thing.
00:15:50.559 --> 00:15:52.480
This is exactly the same thing.
00:15:52.799 --> 00:15:53.279
Interesting.
00:15:53.519 --> 00:15:59.279
Yeah, the we are electrifying so quickly that capacity is being consumed.
00:15:59.360 --> 00:16:03.039
And so I think I think that is one of those high-level themes we see everywhere.
00:16:03.120 --> 00:16:13.120
I don't I haven't I don't know of a single market where they're like, we're sitting around with like a plenty of capacity and like we're in a very fluid area where everybody wants to be, why is nobody doing it's like that doesn't exist.
00:16:13.440 --> 00:16:17.360
So um so I think yeah, capacity is a is a common issue.
00:16:17.519 --> 00:16:39.840
I think what's different is that we see because of the large-scale deployment of renewables across parts of Europe, like affordability is is proving to be very closely connected to um to flexibility, I would say, in terms of if you if you have the ability to use power at certain times a day, you can have very low prices.
00:16:40.000 --> 00:16:44.399
And so that's emerging as like an awareness, I think, more here in Europe than we see in the US.
00:16:44.559 --> 00:16:51.279
Um, just because a lot of our grids just simply don't have as much renewable penetration for the most part, is what we see across some of the markets in in Europe.
00:16:51.519 --> 00:16:54.320
And that's when you see like different use cases for different geographies.
00:16:54.799 --> 00:17:00.240
But yeah, we'll we'll build our learning, machine learning data center here, could be we'll build our cloud, could be.
00:17:00.879 --> 00:17:03.759
Or we just think about storage in ways that the load can be flexible.
00:17:04.240 --> 00:17:06.640
Which brings which brings back to your first question.
00:17:06.799 --> 00:17:08.799
Are we using the grid efficiently?
00:17:09.039 --> 00:17:12.799
And like like you said, uh, I mean, this phenomenon is not new.
00:17:12.880 --> 00:17:16.960
I mean, peak shaving has been used by many industries in the past.
00:17:17.039 --> 00:17:20.799
The data center industry was not using it because they didn't need to.
00:17:20.960 --> 00:17:21.119
Right.
00:17:21.279 --> 00:17:29.039
But now that we're forced to rethink of solutions that already exist and be more efficient about it, peak shaving is amazing.
00:17:29.279 --> 00:17:33.839
It can lower your your your power your power bill, but you need to be smart about it.
00:17:34.640 --> 00:17:40.480
And when you talk about community engagement, again, which will come up on every podcast we talk about here over the next few days, because it's such a big topic now.
00:17:40.559 --> 00:17:49.440
That's how we can help change the perception of the industry is it from the utilization of power to benefit the the community as well as the customer.
00:17:49.759 --> 00:18:03.119
And and that uh again, coming from power, uh, when I got into the data center industry, it was a shock on how little the data center industry does to to break the myths that are created uh along communities.
00:18:03.279 --> 00:18:07.440
I mean, the power industry has been doing that for decades uh successfully.
00:18:07.519 --> 00:18:16.000
Uh and we are, like you said, everyone is still trying to run their own thing instead of joining forces and let's start dismystifying.
00:18:16.160 --> 00:18:21.039
We're not here to increase your power bills, we're not here to drink all your water.
00:18:21.200 --> 00:18:22.480
We are more efficient than that.
00:18:22.640 --> 00:18:25.039
We know what we're doing, let's show that.
00:18:25.200 --> 00:18:31.359
And I think the data center industry needs to be much more vocal and united in doing this.
00:18:31.599 --> 00:18:34.720
And that's that's and that so that community engagement piece is more and more critical.
00:18:34.799 --> 00:18:36.240
More and more projects are being blocked.
00:18:36.319 --> 00:18:39.039
Uh especially now what it's hard in the US at the moment.
00:18:39.200 --> 00:18:39.359
Certainly.
00:18:39.519 --> 00:18:40.000
Certainly, for sure.
00:18:40.079 --> 00:18:46.799
I think I think I think one of the underlying challenges here is that um we are a business-to-business type industry.
00:18:46.960 --> 00:18:47.200
Yes.
00:18:47.359 --> 00:18:50.799
Um, and more than that, we're generally full of engineers.
00:18:50.960 --> 00:19:04.960
And and so like we're probably not we don't have the right DNA in most cases to show up and like listen to a community and talk about things that matter to them because we're, you know, we're gonna show up and talk about how evaporative cooling actually uses less water than a highly electrified data site.
00:19:05.200 --> 00:19:09.759
It's like, no, like people aren't going to understand our business the same way that we do.
00:19:09.920 --> 00:19:17.759
And so I think we just we have not done a great job and we're still trying to make progress to find ways to meet people where they are and have the right conversations.
00:19:17.920 --> 00:19:23.279
And I think that power is one of these places that we can bring real value to the community.
00:19:23.680 --> 00:19:26.400
We can show up and say, we can make your grid more reliable.
00:19:26.559 --> 00:19:32.400
We can actually put downward pressure on pricing for you by putting more power on that existing infrastructure.
00:19:32.559 --> 00:19:40.640
Like we can do these things that impact the kitchen table economics that I think will make an impact um to the community directly.
00:19:40.960 --> 00:20:04.960
And if you go to communities where um where large power uh stations used to lie, whatever type of power, uh, what you'll find is you'll find a willing community that understands economics of big infrastructure and they understand the impact, especially if that a power plant is about to be decommissioned, you have so many families which will will lose their jobs and the multiplying effect disappears.
00:20:05.200 --> 00:20:12.000
So and that's why I was saying the power industry did a good job uh setting up that zone.
00:20:12.240 --> 00:20:13.440
We should use that.
00:20:13.680 --> 00:20:18.160
We should use some of those examples to to relate better with the community.
00:20:18.480 --> 00:20:20.160
It's interesting that you say that.
00:20:20.240 --> 00:20:23.920
I was out at uh Terra Wolf in Buffalo like literally two weeks ago.
00:20:24.000 --> 00:20:29.440
I mean, just uh the scale of what they're building there is just absolutely phenomenal, like absolutely incredible.
00:20:29.680 --> 00:20:36.240
It's really interesting to hear how, first of all, they've taken this old disused coal plant and reconverted it into renewable power.
00:20:36.319 --> 00:20:45.359
And then, second of all, how the community surrounding that power station, like so many people have lost their jobs and they're they're at power distance, so they just went and rehired it.
00:20:45.519 --> 00:20:45.759
Of course.
00:20:45.920 --> 00:20:52.240
I and all of a sudden you hear this kind of you begin to realize that actually there is a really good catalyst with changes.
00:20:52.319 --> 00:20:57.359
And to your point, Chris, like I think um the industry's now increasing mindful of this.
00:20:57.440 --> 00:20:58.480
So I think that's one.
00:20:58.880 --> 00:21:02.160
We're a new industry, it's a really positive step forward.
00:21:02.240 --> 00:21:05.440
But it's certainly like Andy and myself have been talking about this for absolutely ages.
00:21:05.519 --> 00:21:07.119
Like, how do we bring people together?
00:21:07.200 --> 00:21:10.079
It was one of the reasons we founded the DCMC initiative.
00:21:10.400 --> 00:21:25.200
How do you actually bring marketers, public affairs, people, PR, comms people together to actually start addressing these problems of kind of a wider industry level instead of everyone running into the same brick walls in silos, because ultimately everyone's trying to achieve its own goals.
00:21:25.359 --> 00:21:29.920
I think it's about being responsible and being open enough to talk about it and say, we haven't got it figured out, right?
00:21:30.240 --> 00:21:33.519
And you'll find what do you find with communities, which is the most interesting?
00:21:33.680 --> 00:21:38.559
And again, like you said, we're mostly of us are engineers and we completely fail at that.
00:21:38.799 --> 00:21:45.039
But when you start running the exercise, the most interesting thing uh that you'll find is that communities, most of them, they just want to be heard.
00:21:45.200 --> 00:21:45.599
Yeah.
00:21:46.160 --> 00:21:49.359
Listen, literally listen to just want to be heard.
00:21:49.680 --> 00:21:51.039
You've got to listen to the community.
00:21:51.119 --> 00:21:53.519
I think that's a key part of this conversation.
00:21:53.599 --> 00:22:00.000
It's it's not about us telling communities what we're doing, it's about us listening to what their challenges are and solving their challenges.
00:22:00.480 --> 00:22:02.240
Again, that's what the industry is getting much better at.
00:22:02.319 --> 00:22:07.039
I think previously we just go into a community, tell them what we're gonna do, tell them how they're benefit, and they're like, hang on a minute.
00:22:07.839 --> 00:22:22.079
I I actually got told off by uh when I was doing one of these community sessions and say, we're sick of companies who come here, say that they'll do big things, they buy an ambulance, they take some photos, they leave away, and there's nothing here.
00:22:22.400 --> 00:22:25.759
Why don't we need the ambulance, we need other things.
00:22:25.920 --> 00:22:26.559
Yeah.
00:22:27.039 --> 00:22:28.240
It's a it's a good point.
00:22:28.400 --> 00:22:30.400
Less uh less talk, more action.
00:22:30.640 --> 00:22:38.160
A publicity to your point shouldn't be used as kind of a vehicle just to kind of say achieve the objective that you're trying to achieve from a commercial level.
00:22:38.559 --> 00:22:44.960
It actually needs to be responsible and leave something behind so you create, I suppose, a a legacy, the communities in which you're building it.
00:22:45.200 --> 00:22:49.039
And be prepared to have that be something very local, very specific to the whole community.
00:22:49.119 --> 00:22:51.200
There are many communities who might need that ambulance.
00:22:51.279 --> 00:22:53.519
Uh exactly, but but don't assume that everybody does.
00:22:54.480 --> 00:22:56.000
Or at least not just one anyway.
00:22:56.240 --> 00:22:56.400
Exactly.
00:22:56.640 --> 00:22:56.880
Exactly.
00:22:57.359 --> 00:22:59.440
So, guys, we are we are at time.
00:22:59.759 --> 00:23:01.839
Really, really appreciate you taking the time to speak with us.
00:23:02.160 --> 00:23:04.079
I have to ask you the marketing question.
00:23:04.319 --> 00:23:11.200
So uh so just in terms of uh the event, what's really excited you so far and what you're looking forward to?
00:23:11.519 --> 00:23:14.720
Well, it's just getting started, but the uh the award ceremony was last night.
00:23:14.799 --> 00:23:18.720
We won uh Best Global Beta Center Operator Year for the second year in a row.
00:23:18.880 --> 00:23:20.240
So that's been the most exciting thing for us.
00:23:20.559 --> 00:23:21.440
Retained your title.
00:23:21.599 --> 00:23:21.920
Well done.
00:23:22.000 --> 00:23:22.799
Yeah, nice.
00:23:22.960 --> 00:23:24.640
Yeah, multi-award winner now.
00:23:24.799 --> 00:23:25.440
Yeah, yeah.
00:23:25.599 --> 00:23:26.160
Uh yeah.
00:23:26.240 --> 00:23:27.599
So it's been it's been brilliant.
00:23:28.480 --> 00:23:48.480
So it's just started, but it's um I think the the thing that has struck me the most is to see uh first of all how this conference is growing, but also how some some some small companies, and we were talking about one uh a few minutes ago, that uh three, four years ago they were tiny, and now they're massive.
00:23:48.559 --> 00:23:50.240
They're throwing a party in a yard.
00:23:50.480 --> 00:23:58.799
I mean, the industry is growing a lot, and with it it cut you have new new companies coming in, being very successful.
00:23:58.960 --> 00:24:01.200
Uh, and it's great to see them here.
00:24:01.599 --> 00:24:01.839
Right.
00:24:02.000 --> 00:24:08.720
I think events are a great place to realize the scale of that growth because we're all in it, we live it, we breathe it all day, every day.
00:24:08.799 --> 00:24:14.160
You don't really notice it too much, and you come back a year later and you realize there's double the amount of the people, yeah.
00:24:14.240 --> 00:24:17.440
There's another flaw in the in the building, which I found out 10 minutes ago.
00:24:17.519 --> 00:24:22.480
Um, yeah, I think the scale of this industry is a lot off the scale now, hang it.
00:24:22.559 --> 00:24:25.359
And and they're they're already forecasting bigger next year.
00:24:25.599 --> 00:24:26.480
Are they really?
00:24:27.680 --> 00:24:28.720
But enjoy the event.
00:24:28.799 --> 00:24:29.839
Thanks both for coming on.
00:24:30.079 --> 00:24:30.240
Thank you.
00:24:30.400 --> 00:24:34.000
Thanks for the that's it for this episode of the Inside Data Centre Podcast.
00:24:34.079 --> 00:24:35.039
Thanks for tuning in.
00:24:35.200 --> 00:24:43.119
This show is brought to you by Data X and Data Exec, the recruitment partners powering global talent across the data centre and digital infrastructure world.
00:24:43.359 --> 00:24:47.200
If you enjoyed today's conversations, don't forget to follow, like, and subscribe.
00:24:47.359 --> 00:24:49.839
It really helps us grow the podcast community.
00:24:50.079 --> 00:24:55.440
Stay tuned for more stories, insights, and conversations with the people shaping the future of our industry.
