All Episodes

July 25, 2025 32 mins

The rise of AI is pushing data centers to their limits. Calls for bigger, denser facilities are increasing and tenant power requirements are doubling in many markets.

It’s still not enough to meet the needs of today’s market or tomorrow’s users.

On this episode, Bisnow National Data Center Reporter Dan Rabb said keeping the momentum of the data center boom is all about finding power. And while possible solutions abound — microgrids and nuclear among them — being provided today aren’t going to come through fast enough.

Register on Bisnow.com to join next Friday's conversation live, or check back here for the conversation after it airs. 

Mark as Played
Transcript

Episode Transcript

Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Mark Bonner (00:10):
Okay. Welcome to First Draft Live. It's Friday,
July 25. I'm your host, MarkBonner, BizNell's editor in
chief, I'm coming to you livefrom New York. Thank you.
Thank you to so many of youtuning in from across The US and
around the world. On today'sepisode, can the grid keep up or
is the data center surge aboutto break? AI demand has

(00:32):
supercharged leasing and rents,but the infrastructure cannot
keep pace and the cracks arestarting to show. Transformers
are backlogged, transmissionlines are stuck in permitting,
and in some markets, developersare waiting up to five years,
five years, Dan, just to plugin. Now PJM, the country's grid

(00:53):
operator, country's largest gridoperator, is under political
fire.
And from Phoenix to NorthernVirginia, fully funded sites are
frozen, tangled in delays,politics and grid dysfunction.
Our guest today, my colleague,Dan Rab, Bisnow's national data
center reporter. Dan's beentalking to developers, power

(01:15):
providers, and hyperscaletenants for years. And today
he's here to help us make senseof the cracks forming an
industry's hottest hottest assetclass. We'll also get into the
capital stack and why datacenter CRE is starting to look
more like infrastructurefinance, very annoying, very
interesting development in thelast couple of weeks.
Also to the audience, we'retaking questions from you today,

(01:39):
AMA style. Send them into thechat now and my producers will
get this to me as soon aspossible. Dan, welcome to the
show. I know you're calling fromBoston.

Dan Rabb (01:49):
Yeah. Thanks for having me on and for the kind
introduction. Happy Friday andexcited to be chatting about
data centers with you.

Mark Bonner (01:56):
Okay, let's jump on in. Nearly 80% of U. S. Data
center demand is concentrated injust six markets, and all of
them are power constrained asper your reporting. In PJM, home
to Northern Virginia and Ohio,average interconnection wait
times now exceed five years.
Lead times for key componentslike transformers and switch

(02:17):
gears are eighteen to thirty sixmonths. And what's going on
here? What's the latest?

Dan Rabb (02:25):
Well, you know, overall in the data center
industry, you know, power is thelong pole in the tent when it
comes to development, thescarcity of power, you might
say. It's the biggest challengethat developers are solving for
right now, among many, right?It's not we hear about these
waves of demand, you know, andthe growth of the industry, but

(02:46):
it, you know, there's laborconstraints, severe labor
constraints that have slowedgrowth. There's supply chain
constraints. You werementioning, you know, wait times
for transformers.
It really is, though, power thatis kind of superseding all those
things. And just to give you asense of it, and I don't wanna
oversimplify things here, but aneasy way to think about it is if

(03:08):
you turn back the clock justfive years, seven years, you
know, a data center that had a afew dozen megawatts of capacity,
you know, 30 megawatts, that wasconsidered a pretty big data
center project. Today, we'relooking at projects that are a
few 100 megawatts beingannounced regularly. And so say

(03:31):
you're trying to build, it'sfive to seven years ago, you
wanna build a 50 megawatt datacenter in Northern Virginia, the
world's largest data centermarket.

Mark Bonner (03:40):
Loudoun County.

Dan Rabb (03:40):
Yeah, Loudoun County. Five to seven years ago, you get
some land near transmissionlines, you get in touch with the
utility, they give you a lettersaying you'll have power by a
certain time. And we're talkingprobably months, not years,
right? And the amount of timevaries depending whether they

(04:00):
need to build a substation,things like that. But you could
count on, know, they saidthey'll have power to you by a
certain date.
There's no mystery there. You'regonna get power by that date.
Fast forward to today, thatsituation looks really, really
different. You know, there'smore data centers being built,
and the scale of those projectsis just way, way bigger. You

(04:23):
know?
Campuses with multiple datacenters that have fifty, sixty,
70 megawatts each, you know,it's sort of wild to see how
it's gone from, you know,there's megawatt scale campuses
being built right now. Know,that was something that we wrote
an article two years ago saying,are we ever gonna see a gigawatt
data center campus? And now, youknow, that's at least planning

(04:47):
for those is the new normal. Sonow, if you wanna build say a
500 megawatt data center inNorthern Virginia and any of the
traditional hotspots, you go tothe utility, you say you need
500 megawatts, it's a lotdifferent situation than it was
five to seven years ago. Firstof all, you're one of 50
developers who's asking for thatsame sort of thing.

(05:08):
And the utility, who's legallyobligated to fulfill those
requests is looking at thisqueue of gigawatts of power that
people are asking for. It's way,way more than they can possibly
fill, right? It's overwhelmedwhat the grid infrastructure can
carry. A transmission line canonly carry so much power. Right?

(05:28):
A power plant can only produceso much power. So now it's not
just them building like asubstation for you and running a
little transmission line toredeveloping an entire
transmission corridor, know, orit's building new power plants.
These are things that take, youknow, years to do, take billions

(05:49):
of dollars to do. So they may,and so now they're saying, well,
it's not six months, it's sevenyears to get power, but also
that seven year timeline isreally kind of an estimate,
right? You can't really count onthat.
So the hyperscalers who aredriving all this demand, you
know, the big tech companies,the Googles, Microsofts, you

(06:11):
know, data center procurementteams who are driving demand
here, they have desks full ofpeople saying, hey, we're gonna
have, you know, 300 megawattsfor you at this site in this
market by this date, but theydon't really trust any of that.
And so basically the winnershere are going to be the
companies that can reliably getpower quickly for these

Mark Bonner (06:35):
ideas,

Dan Rabb (06:36):
if that makes sense.

Mark Bonner (06:36):
And this situation has gotten so dire in recent
months that it's becomepolitical. And now there's real
questions nationally about who'sactually in charge of the grid
in The United States. PJMinterconnection, which serves
about 65,000,000 people give ortake, is under political fire.
Nine governors signed a letterthis month accusing PJM of
dysfunction and threatening toleave. Meanwhile, FERC, that's

(06:59):
the federal regulator in chargeof transmission, is also in
leadership transmission, whichis throwing a little bit more
chaos into this conversation.
Yet PJM at this moment controlsnearly 30% of US data center
development. Dan, I know you'vedug into this tension. Does
anyone actually have the power,no pun intended, to speed things

(07:20):
up?

Dan Rabb (07:21):
Yeah. You know, it's it's a good question and a and a
timely one. I mean, I guess thewhere I'd start with this is
just the way that our power gridand electric transmission
infrastructure is structured inThe US is just incredibly
labyrinthian and balkanized andmultilayered in a way that makes

(07:44):
any sort broad scale reform verydifficult. And just to kind of
outline it, you you mentionedPJM. PJM is one of seven kind of
regional, they're called RTOs orISOs, but they're these regional
transmission organizations thatbasically run the grid in
sections of the country,separate grids.

(08:05):
And within these organizationsare the utilities who actually
sell power to customers. Andlet's take PJM as an example.
PJM has 1,200 member utilities.Right? So over a thousand member
utilities underneath it who allhave their own operations and
ways of doing things andinteractions with customers.

(08:27):
You also have each state within,you know, has their own utility
commissions and their own lawsthat regulate power
transmission. So that's anothersort of actor and complicating
factor within that. And then ontop of that, you have sort of
the federal regulatory bodieslike the Federal Energy
Regulatory Commission, FERC, whosort of oversee all these

(08:50):
transmission organizations. Soyou think about all these
different bodies with their ownagendas and their own, who all
kind of are a different piece ofthe puzzle, you know? And the
amount of stakeholders you haveto get on board with any kind of
change is really significant.
And so it's just not a systemthat's particularly well suited

(09:12):
to making the widespread changesthat a lot of people I talk to
say are needed right now ifwe're going to, you know, if the
demand from data centers isgoing to be fulfilled, if that
makes sense.

Mark Bonner (09:29):
Right. And like in this broadly in commercial real
estate, you know, the housingcrisis here is a great example
of this. There's a whole lot ofenergy around interest rates is
another conversation about thiswhere it's like things that the
industry says it needs yesterdaybut then can't come to pass for
years ahead. Housing crisiscan't be solved in the next ten
years. It's going to take a longtime.
Interest rates were still waitand see mode. We'll see what

(09:50):
happens with Powell next week.And then this is another one
where even in my own sourcesthat I talked to here in New
York at our last data centerevent, the conversation isn't
about can I get the capital tobuild a data center? The
question is, how am I going toget the power to power the data
center once I build it? Buildingit is one thing, powering it is

(10:12):
another.
So I guess along those lines,Dan, in your opinion, just based
on your reporting, what is thebest case scenario for a reform
timeline as we're looking at TheU. S. Grid?

Dan Rabb (10:23):
Yeah, no, it's a good question. And, you know, it's I
think there's disagreement aboutwhat we're talking about when we
talk about, you know, reformsand the changes that are needed.
You know, there's this thetransmission organizations like
PJM and utilities are dealingwith a pretty difficult

(10:44):
situation. Right? These arepower demand in The United
States was pretty much stagnantfor decades.
These are not organizations thatare used to existing in an
environment in which there'sthis surge of growth that they
need to somehow fulfill. They'realso dealing with this surge in
demand while there is atransition towards renewable

(11:10):
energy. Right? Most of theprojects that are in what they
call the interconnection queue,which is sort of the line
waiting to be connected to thegrid, you know, are solar
largely solar, also wind batteryprojects, things like that as
well. And those behave verydifferently from, you know,

(11:30):
fossil fuel sources, likenatural gas coal, that kind of
thing.
And so it's been a hugechallenge for them to sort of
figure out how to ensurereliability of the grid, ensure
that costs stay down, and thatwas even kind of before you
added the surge of demand intothe picture. So there is a lot
of disagreement though, between,you know, say the Trump

(11:54):
administration, data centerdevelopers, activists or
advocates for clean energyabout, you know, not just what
reforms are needed, but likewhat the end goals of those
reforms are. Are we just tryingto meet demand for data centers?
Even if that means, you know,saying, okay, all you solar

(12:15):
projects, you wait to the sidefor years. We're just gonna hook
a lot of natural gas.
You know, what are the end goalsthere? And so there's
disagreement even at that level.So, you know, I'm not hearing a
lot from folks talking aboutspecific timelines yet. It's
almost just trying to get theball rolling, if that makes
sense.

Mark Bonner (12:33):
Makes perfect sense. And you brought up Donald
Trump. Let's talk about it.Earlier this week, Donald Trump
unveiled his AI action plan. Itexplicitly encourages self
powered campuses.
Meta is investing in on-sitenatural gas generation in Ohio
and Louisiana. Amazon isexploring modular microgrids.
The logic is if utilities can'tdeliver, then the industry needs

(12:57):
to build their own. As you know,more than six gigawatts of
privately operated generation isalready planned or under
construction throughout thecountry and that number would
triple by 2027. Dan, in yourreporting on this this week,
what are your sources tellingyou about whether or not this
workaround could actually becomethe new normal?

Dan Rabb (13:19):
It's a good question. And, you know, it's definitely
something we're seeing more of,right? Data centers, really data
center campuses being builteither with some form of power
generation, almost alwaysnatural gas, you know, either on
the campus or contracteddirectly with a, you know,

(13:40):
energy developer as opposed togoing through the grid, turns
behind the meter in a in theindustry jargon. And and we're
definitely seeing more of that.Right?
It's something you really sawnone of, you know, a a year or
two ago, and now a lot of,there's a lot of projects being
pursued with that sort of thing.As far as whether it becomes the

(14:01):
new normal is a differentquestion. It's really easy to
see the appeal of it, right? Youcan't get grid energy, why not
just produce our own? A coupleof things stand in the way of
really widespread adoption.
And just to name a couple ofthem, first of all, the large

(14:25):
tech companies who are by farthe largest data center users
are the, they're the tenants whopeople are trying to bring on or
get powered land so that theywill build on that land. They
still view utility power as likethe gold standard, right? For
cost reasons, really forreliability reasons, they will

(14:45):
at least having heard from acouple of sources at companies
like Microsoft, they'll wait alittle longer for a project that
will give them power from thegrid as opposed to getting power
from maybe on-site generation orbehind the meter natural gas,
even if that's a little bitfaster. Again, reliability and

(15:07):
cost being big reasons there.It's also, while it's easy to
think about it as sort of aworkaround, these projects also
face their own sort of uniqueconstraints.
So just as power transmissioninfrastructure is constrained,
right, so too is natural gasinfrastructure, right? Like

(15:27):
these, it's not a limitlesssupply of natural gas flowing
through a gas pipeline. So theability of a particular data
center site to have as much gasas it needs to fuel itself, but,
you know, to produce that powerall the time is not always so
clear. And it's not clear thatwill be a limiting factor in how

(15:49):
many of these projects get builtout. And then the final thing
I'd say is a lot of theseprojects require or basically
ask the data center firm to haveexpertise in an area where
they're really experts.
Know, these are companies thathave, you know, not historically
done their own, you know,on-site generation or even

(16:11):
understood sort of some of theunderlying issues. It's a very
steep learning curve. And Ithink a lot of companies are
just for good reason,uncomfortable sort of, going
outside of their, well, outsideof their comfort zone, so to
speak.

Mark Bonner (16:27):
If you're just tuning in, we're talking with
Dan Rabb, Business National DataCenter reporter, about how the
AI Gold Rush colliding with avery analog problem. The grid
can't keep up. Dan, let's go tosome questions from the
audience. Microsoft hascommitted to hellion for power
from the first power positivefusion generator. What does the

(16:48):
rest of the industry think aboutthis tech?
What are your sources tellingyou?

Dan Rabb (16:52):
You know, it's interesting. So fusion, and I
will admit, I'm notoverwhelmingly familiar with the
specific deal that you'retalking about, but I will say
fusion in general, I think it'sviewed I'm gonna quote somebody
from one of our BizNow datacenter conferences where they
said fusion is the technology ofthe future and it always will

(17:14):
be. That is just one perspectiveI've heard, but just basically
what I the sense I get is thatthere are power issues right
now, today that need to be dealtwith. And so there are solutions
that are out there maybe yearsin the future that there's a lot
of intrigue around and interestin. But the focus is really on

(17:39):
things that can have an answerin the next year, two years,
three years.
And, you know, I'll extend thatout to, you know, we're talking
about nuclear, you know, smallnuclear reactors, right? There's
a ton of buzz around what theycall SMRs in the data center
sector. These are, you know, afraction of the size of a think

(18:00):
more of like the nuclearreactors that power an aircraft
carrier than, you know, your 3Mile Island style nuclear power
plant. And there's a bunch ofcompanies trying to bring these
SMRs to market. It's thetechnology isn't there.

(18:21):
It's making it, it's bringing itto a price point where you can
make it marketable and where itcan be competitive. So that is
something that I've talked abouta lot, but in the best case
scenario, what I hear is that,you know, it's five years out
before the first of these SMRsare deployed and, you know, the

(18:42):
ability to scale it up, youknow, to use it in a more
widespread way is, you know, tenyears out minimum. So, you know,
again, a lot of attention paidto these things, but think it's
sort of being pushed to the backof people's minds because
there's such an immediate needto address these power concerns.

Mark Bonner (19:02):
Right. And that's just another cost concern. And
that's another timeline that'sway out into the future where
there seems to be this mad dashto getting things done now, but
at every level of the governmentand in every level of finance,
right? Case in point, tariffs.The US now has an average
effective tariff of 15.5%,construction materials, servers,

(19:24):
power infrastructure, they'reall affected.
Developers report a 1,000,000 to1,200,000.0 cost increase per
megawatt just from Trump's tradepolicy. Meanwhile, supply chain
volatility continues, lead timesare rough, and tariffs on
Chinese steel and electricalgear remain in flux. How does
that factor into the equationwhen we're talking about this

(19:48):
other issue with power? And thennow you layer on economic policy
out of the White House on top ofit.

Dan Rabb (19:54):
Yeah, it's interesting. In some ways, the
power constraints end up beingalmost a saving grace because if
that's your long pole in thetent, then, you know, if the
supply chain constraints can,you know, they're not pushing
project deadlines back justbecause there's this other
factor that is pushing theproject deadline back so much.
But that said, you know, almostuniversally, you know, what I've

(20:19):
heard from folks in the industryis that, you know, the tariff
situation does create ameaningful impact and is making
an already difficult supplychain situation a little more
difficult to navigate. You know,most folks recall, you know,

(20:40):
during the pandemic, globalsupply chains got scrambled
quite a bit. For a variety ofreasons, the data center sector
has never really come out ofthat, right?
You're still, for crucial items,transformers, things like that,
down to basic buildingmaterials, You know, lead times

(21:00):
of, you know, a year or moreapproaching two years, it shifts
pretty rapidly, but you're stilllooking at two years for a
generator and a transformer andthings like that. And, and this,
and the tariff situation has,you know, people were already
pursuing really creativesolutions to try to make sure
that, you know, when you getpower and it's go time, let's
build that you actually haveeverything you need on-site to

(21:23):
get the project done. Andcertainly tariffs have made that
more difficult. You know, ithas, I do hear from some folks
that there has been the intendedimpact of the tariffs and a lot
of companies looking to onshoresome degree of manufacturing if
possible. You know, obviouslythe amount of, it takes a long

(21:46):
time to do that sort of thing,but it is something that a lot
of

Mark Bonner (21:52):
I supply chain are looking mean, there seems to be
a lot of shifting dynamics here,right? The Trump administration
wants the near shore, onshorethings into The United States,
right? But then we have thistariff situation. We have these
costs that are going up. Itmight actually be cheaper to do
these things overseas.
Does a data center have to be inThe United States? Could it be
in Asia? Could it be in Europe?Could it be in South America?

(22:12):
The Trump administration made astrategic decision to allow
Nvidia to sell its ships toAsia, China directly.
Right? Are you hearing from yoursources that maybe site
selection is going to shift moreaggressively overseas, or is
this just more negotiation painvia the White House?

Dan Rabb (22:32):
So the answer is yes, but it's important not to
overstate the case. Just by wayof background here, a lot of
data center site selection hasto do with the particular
computing workload that is goingto be housed inside that data
center. Like it's the reason whydata centers tend to be
clustered together in certainmarkets. Right? The term is

(22:55):
latency.
You know, the amount of time ittakes data to flow from the data
center to where the end user isand back again. You know, so
it's unlikely, for example, thata data center that is being
planned for Northern Virginia,which tend to be the most
latency sensitive, it's therefor a reason. You know, that's

(23:17):
not just gonna, they're notgonna say, oh, it's too
expensive to build here. Let'sbuild this in Japan, right? Or
South America.
But that said, there are usecases that are more location
agnostic, you know, particularlywith AI, you know, certain AI
workloads are less latencysensitive. And you can extend

(23:38):
that out to, you know, somecloud providers, you know, today
have data centers on the EastCoast and on the West Coast that
serve, you know, some marketsfor certain types of computing,
you know, in Asia Pacific or inEurope for certain things. And
so around the margins there withboth of those cases, you know,

(24:00):
you could see situations where,look, today The United States is
by far kind of the cheapest andeasiest place to build data
centers. If that changes, whichmaybe tariffs do change that,
Yeah, these use cases around theedges could shift, but it's

(24:24):
around the edges. Like I said,you're not going to see things
that were meant for these coremarkets suddenly going
elsewhere.
But AI, some of that for sure,possibly.

Mark Bonner (24:34):
Let's get another question from the audience. Dan,
it was announced this week thatthere would be support on the
federal level for developingdata centers of 100 megawatts or
larger on federal land. Giventhe recent pushback in local
communities to scale down thesize of data center projects in
their respective areas, do yousee a bottleneck forming in the

(24:55):
push to develop on federal land?

Dan Rabb (24:57):
It's a really interesting question. And yeah,
there was a, there was a just togive a little background here.
So there was an executive ordersigned by president Trump on
Wednesday that, among otherthings, it encourages data
center development anddevelopment of energy
infrastructure for data centerson federal land. It should be

(25:19):
noted this continues a policy ofthe Biden administration.
President Biden also signed anexecutive order doing that exact
same thing.
There are some differencesbetween the two surrounding
labor and hiring practices and afew other things and focus on
renewable energy and things likethat. But the idea of building
on federal land is somethingthat is has had bipartisan

(25:44):
support behind it.

Mark Bonner (25:46):
Dan, when we talk about federal land, what kind of
land is this? I mean, this isn'tlike national parks, is it? I
mean, what are we talking abouthere?

Dan Rabb (25:53):
No, know, it's particularly in the Western U.
S, there's a huge amount offederal land, you know, Bureau
of Land Management. And I don'twanna get too far out over my
skis when, you know, in terms ofwhat I know about the specific
types of, parcels they'd belooking. I know some of the

(26:13):
specific policy documents fromboth administrations had talked
a bit about sites near oradjacent to national
laboratories, Idaho NationalLaboratory, things like that.
But as far as specific sites onfederal land, I don't have a ton
of insight into that,unfortunately.

(26:34):
I wish I did. But the questionof pushback is a really
important one. You know, I wouldsay of the challenges facing the
industry right now, one of thebiggest is the rise of
opposition at the local level,right? Until very recently,
these were buildings that weremost of all didn't know what

(26:56):
they were, right? And they wereonly built in areas with a lot
of industrial, things like that.
Now, as power gets more scarce,they're increasingly going to
wherever they can find powernear these major markets. And
that means near residentialareas or into, you know, more
rural communities that aren'tused to having somebody build a
giant industrial structurethere. Right. And don't like

(27:17):
what that does to

Mark Bonner (27:17):
their time. NIMBYism. They don't create a
lot of jobs. They don't producea lot of taxes for the
community. Yeah, you've writtenabout this for years, Dan, and
that's still ongoing.
And it's because a lot of thesesmall communities around the
country over the five or sixyears have gotten wiser about
what data centers are and whatthey are not. And just because
you see a big company like Metaor Google come in which that

(27:41):
could get you excited about,right? It's not exactly what you
think it is going to be once thedata center is completed. Is
that correct?

Dan Rabb (27:50):
Yeah. And you know, look, think it's a very under
Regardless of where one falls onthese issues, most people don't
want a giant concrete structurebuilt across the street from
their house, right? It's anunderstandable perspective,
know, at least I believe. But,you know, there's been a growing

(28:11):
amount of kind of organizedopposition. And like you said,
understanding of some of thebroader issues, the potential
impact on like rising powerprices, things like that.
You know, there's some stuffthat is true, some stuff that
maybe there would be a lot ofpushback to from a factual
perspective, you know, maybearound water use and things like
that. But the difference betweentoday and five years ago is that

(28:32):
pretty much everybody knows whatdata centers are. Everybody's
read stories in their localpaper about, you know, the
person's well, stop producingwater because the data center
right there. And the industry,the data center developers have
had to adjust to this. They havenot been used to having to fight
these battles at the locallevel.
And it's been a big change and abig adjustment and a big

(28:54):
challenge that they're just sortof getting their heads around
right now. And it's only goingto be more of an issue going
forward. And, you know, toanswer the listener's question,
in terms of how that impactsfederal lands and the move
towards that, I mean, I thinkmaybe the idea is that federal

(29:15):
lands will have less of anability to have that sort of
local pushback. Right? But I'mnot sure it's something that's
very much worth, looking into.

Mark Bonner (29:24):
Okay. My producer's furiously telling me to wrap it
up, but I do want to go to onemore question from our audience.
So it's kind of an interestingquestion for a reporter. So feel
free to pass on this, but inyour reporting, Dan, given the
economic and politicalconstraints, are hyperscalers
expectations unrealistic giventhe moment? What are you
hearing?

Dan Rabb (29:45):
So expectations in terms of how much they want to
build out, like the amount ofcapacity they want to build?

Mark Bonner (29:54):
I'm going interpret this as the difference between
what they can get financed andbuilt versus what can actually
come online. Right. I mean, weknow the demand for data centers
is we haven't met the demandyet. Right. We're still trying
to meet the demand.
Right. And this is one of thefastest growing things of the
last ten years in the commercialreal estate space. I think
that's the question. Right.

Dan Rabb (30:17):
Yeah. So I I think it's a it's a really good
question, actually. You know, Ias a journalist, I'm sort of
trained to not project my ownopinions out there. What what I
hear from folks, I don't hearfolks saying this is
unrealistic. You know, that it'sthat, you know, the planned
CapEx or whatever they saytheir, you know, runway is for

(30:42):
building out their computingcapacity that, you know, that
sort of market conditions andsome of these constraints that
we've been talking about willnot allow that to happen at the
pace they want.
You know, I think there arequestions, serious questions
around, you know, when we lookat some of these big projections
of data center demand that youhear from utilities, you know, X
number of gigawatts in thedevelopment pipeline, things

(31:04):
like that. A lot of thatprobably won't materialize. You
know, a lot of Microsoft, forexample, I'm just using them as
an example. I'm not saying themspecifically, but might plan
three projects in one market andthen only build one, the one
that gets power first. Right?
So some of these pipelines forcertain areas probably will not
end up being as big as they are.I think there's widespread

(31:27):
consensus on on that.

Mark Bonner (31:31):
Okay. That's all the time we have for today. Dan,
thanks so much for being here.

Dan Rabb (31:35):
That was a lot of fun. Thanks for having me.

Mark Bonner (31:37):
Alright. We'll be back with another episode of
First Draft Live next week, sodon't miss out. You can sign up
now on our event page. You canalso find today's episode and
all of our past conversations onyour favorite podcast app. I'm
partial to Spotify, Dan.
You can also check us out onbiznow.com. This is First Draft
Live. Have a great weekend,everyone.
Advertise With Us

Popular Podcasts

New Heights with Jason & Travis Kelce

New Heights with Jason & Travis Kelce

Football’s funniest family duo — Jason Kelce of the Philadelphia Eagles and Travis Kelce of the Kansas City Chiefs — team up to provide next-level access to life in the league as it unfolds. The two brothers and Super Bowl champions drop weekly insights about the weekly slate of games and share their INSIDE perspectives on trending NFL news and sports headlines. They also endlessly rag on each other as brothers do, chat the latest in pop culture and welcome some very popular and well-known friends to chat with them. Check out new episodes every Wednesday. Follow New Heights on the Wondery App, YouTube or wherever you get your podcasts. You can listen to new episodes early and ad-free, and get exclusive content on Wondery+. Join Wondery+ in the Wondery App, Apple Podcasts or Spotify. And join our new membership for a unique fan experience by going to the New Heights YouTube channel now!

The Breakfast Club

The Breakfast Club

The World's Most Dangerous Morning Show, The Breakfast Club, With DJ Envy, Jess Hilarious, And Charlamagne Tha God!

Fudd Around And Find Out

Fudd Around And Find Out

UConn basketball star Azzi Fudd brings her championship swag to iHeart Women’s Sports with Fudd Around and Find Out, a weekly podcast that takes fans along for the ride as Azzi spends her final year of college trying to reclaim the National Championship and prepare to be a first round WNBA draft pick. Ever wonder what it’s like to be a world-class athlete in the public spotlight while still managing schoolwork, friendships and family time? It’s time to Fudd Around and Find Out!

Music, radio and podcasts, all free. Listen online or download the iHeart App.

Connect

© 2025 iHeartMedia, Inc.