WEBVTT

1
00:00:00.060 --> 00:00:03.140
On y'all. Welcome back to Blockspace Live, presented by CleanSpark.

2
00:00:03.260 --> 00:00:20.360
We are rolling through earnings week, week two, and today we've got two banger interviews with, first, Sam Tabar of White Fiber, CEO of White Fiber, to discuss their Q2 earnings, some questions we have on their blended cloud and powered shell business segments.

3
00:00:20.960 --> 00:00:29.300
Following that, we will be tackling Iron delivering the first tranche of 50 megawatts to Microsoft at its Childress facility in Texas.

4
00:00:29.940 --> 00:00:47.420
And then we have none other than Captain Keel himself, Ben Gagnon, CEO of Keel, to discuss their Q2 earnings and what it's going to take to get their sites in Pennsylvania across the finish line for their first tenant, and also why they decided to decline an expansion for 10 megawatts at their site in Washington.

5
00:00:48.080 --> 00:00:51.470
To round off the show, [smacks lips] we've got a feel-good story for y'all.

6
00:00:51.940 --> 00:01:01.000
Specifically, how data centers are bringing additional tax revenue and helping to enrich public services at the localities in which they are based.

7
00:01:01.620 --> 00:01:15.060
Little data center propaganda for y'all today, but we have to look at the facts on the ground as to how these are actually helping communities, given how much vociferous pushback there has been all across the US for building in certain places.

8
00:01:16.660 --> 00:01:28.680
That's right. Blockspace goes live every weekday at 1:00 PM Eastern. We are compute's daily live show featuring quick hits on AI, data centers, emerging technology, and markets.

9
00:01:29.120 --> 00:01:35.650
If you miss the live stream, it turns into a podcast shortly thereafter. You can find that anywhere podcasts are found. And then

10
00:01:36.560 --> 00:01:48.360
if you still can't get enough of the podcast and live stream content, we do a lot of written stuff. You can find us on, on Yahoo Finance, but the best place to go is our website, blockspace.media.

11
00:01:48.490 --> 00:02:05.160
Again, our website, blockspace.media. This live stream is brought to you by CleanSpark, NASDAQ listed ticker CLSK. More on them later on in the show. And we are on day four of week two of earnings week two.

12
00:02:05.660 --> 00:02:21.480
Play the intro. [upbeat music] Yeah, uh, we're not gonna do an earnings update today 'cause we have back-to-back, uh, CEO interviews with, uh, your, the who's who of the Neocloud and Powershell game.

13
00:02:21.820 --> 00:02:28.680
But before that, I wanna show, um, that here's a little teaser of what we're, uh, what, what the week's looked like so far.

14
00:02:29.320 --> 00:02:44.960
Um, today we have Ben of Keel, Sam of, uh, White Fiber, and they were doing an earnings stream of Soluna at the end of the day. And tomorrow we have Soluna CFO, Mike Pichey to close out the earnings week coverage.

15
00:02:44.980 --> 00:02:51.640
For those of you who've been paying attention, thank you very much. We've enjoyed it. Tell us what you do and don't want to see next earnings.

16
00:02:51.970 --> 00:03:05.960
Again, this is roughly every quarter, so three months, we'll just run it back. Uh, that's what I have on the earnings, Colin. Um, I think if you're ready, we can bring up Sam. Yeah, let's get Sam up here.

17
00:03:06.180 --> 00:03:15.980
Yeah, he's in the wings. He's dressed for today with that sports coat. Oh my goodness, looking sharp. Sam, welcome to the show. Yeah, I just wanted to look good for you guys, you know. [laughs] Yeah.

18
00:03:16.580 --> 00:03:25.660
Every day is earnings day. Uh, well, [claps hands] so we've got you for a little bit here. Um, I'll just jump in and, uh, throw the first question out, which is,

19
00:03:26.680 --> 00:03:40.900
so, uh, you had a customer pay $12.3 million to walk away from the deal, and that's in the light of this being one of the tightest GPU markets we've ever seen. Um, that's like the only thing people care about.

20
00:03:41.720 --> 00:03:52.620
Uh, why was this deal terminated, or did you guys push to terminate? Did the tenant push to terminate? Um, yeah, I'm kinda curious, like, behind this termination here.

21
00:03:52.980 --> 00:04:02.680
I mean, well, first of all, we're, we're not unhappy about it. Uh, we were able to recognize $12.3 million of, uh, termination-related revenue, uh, in the second quarter.

22
00:04:02.820 --> 00:04:12.920
The total termination payment is approximately $15.7 million, with the remaining amount, um, expected to be collected in subsequent quarters.

23
00:04:12.980 --> 00:04:18.980
The benefit to profitability, uh, was partially offset by the associated equipment lease.

24
00:04:19.560 --> 00:04:28.620
We also experienced some downtime, uh, while transitioning that capacity to replacement customers, and those replacement contracts were signed at lower rates than the original agreement.

25
00:04:29.040 --> 00:04:33.120
But that's exactly why pr- contractual protections matter.

26
00:04:33.640 --> 00:04:44.020
The termination, [car engine revs] uh, payment helped bridge the economic difference and protect the value of the original engagement while we re- r- pl- while we redeployed the equipment.

27
00:04:44.080 --> 00:04:55.500
So we were able to turn a customer termination into an orderly transition rather than a permanent loss of value. And more importantly, we have moved well beyond that transition.

28
00:04:55.540 --> 00:05:03.200
Since our last earnings call, we have signed more than $540 million of new multi-year cloud services agreements.

29
00:05:03.240 --> 00:05:17.400
So as those deployments begin entering later this year, we expect them to drive substantial growth in rec- recurring cloud revenue. Uh, so Sam, kind of just a quick follow-up question there for the specifics.

30
00:05:17.420 --> 00:05:28.360
The replacement, is that inside the $540 million cloud revenue, as- Partially, yeah... discussed on the 10-Q? Partially. And, and the, the revenue has already commenced, correct? Yeah, some of it. So our...

31
00:05:28.560 --> 00:05:40.940
Yes, def- I believe so, yes. The, the $540 million, some of that I believe has already commenced, but there, it, that's gonna, a lot of it is taking place over the next few months.

32
00:05:41.000 --> 00:05:49.180
Our cloud pipeline is just surprisingly very strong. Um, several of these agreements were just signed over the past several weeks.

33
00:05:49.480 --> 00:06:03.860
Uh, we're seeing real commercial momentum, um, as larger customers consolidate deployments with providers that can source advanced hardware, execute complex deployments, and operate them reliably.

34
00:06:04.480 --> 00:06:18.880
Uh, you know, these include Nvidia B300 deployments in Ontario, uh- Even Vera Rubin deployments, uh, we've been able to, uh, do, and that's- there are not many of our peers that can say that they've been doing Vera Rubin deployments.

35
00:06:18.940 --> 00:06:31.120
So these are, these are high-quality counterparties. We're doing very high-quality sophisticated work. Um, and, uh, I'm just really proud of the momentum and the reputation that our cloud business is, is doing.

36
00:06:31.440 --> 00:06:41.720
And I should just reiterate that our cloud business and our colocation business are two different teams, and, uh, they're just-- they're two different operations. There's no crossover.

37
00:06:41.800 --> 00:06:54.440
There is gonna be some potential crossover in the near future, but those are two different revenue line items for White Fiber. And, um, 'cause you brought up Vera Rubin, I'll-- let's jump straight to a GPU question.

38
00:06:54.520 --> 00:07:13.220
So, um, it looks like Prime Intellect's Rubin deal, if I'm doing the math right, comes out to roughly, like, $7 per GPU hour, and Morgan Stanley says it costs, what, $7.8 million to build one of those racks, which is about double what the Blackwell hardware costs, so it's a bit of a price premium.

39
00:07:13.820 --> 00:07:26.500
Like, what do you think about Rubin, uh, it just in terms of, like, CapEx per dollar versus Blackwell? Also, like, how did you guys get the NVIDIA deal? Like, one of the- the first small cap Rubin allocation.

40
00:07:26.540 --> 00:07:35.600
I'm kind of-- Yeah, what's going on there? Yeah, listen, I'd like to focus on, on that, especially that [coughs] excuse me, that second part of the question is really important.

41
00:07:35.640 --> 00:07:43.160
We have really-- We have very deep and excellent relationships with NVIDIA. They've been our partner for a very long time.

42
00:07:43.200 --> 00:08:03.259
In fact, th-that relationship occurred, um, th-this is a little-known fact, but NVIDIA's revenues, once upon a time, over ten percent of it was because they were building, uh, ETH mining machines when Ethereum was under proof of work, before it shifted over to proof of stake.

43
00:08:03.760 --> 00:08:09.400
And those relationships during those a- now ancient days continue to carry over.

44
00:08:09.520 --> 00:08:21.240
Uh, we were one of the only Bitcoin miners once, when we were a Bitcoin miner, to actually do ETH mining, and so that, that r- those relationships with NVIDIA carry over from, from many, many years.

45
00:08:21.800 --> 00:08:27.820
And so our relationships go way back. They go very deep. And you're right, it's not very common.

46
00:08:27.860 --> 00:08:44.260
In fact, I can't think of any other peer that was able to get their hands on Vera Rubin, and that's a testament to our relationship with NVIDIA. And yeah, so I'm curious, uh, just the, the GPU economics.

47
00:08:44.420 --> 00:08:56.120
I mean, uh, the, the Rubins for Prime Intellect, uh, just higher CapEx, a premium. What do you think? I mean, how do you think about GPU economics and signing those deals?

48
00:08:56.140 --> 00:09:09.070
Well, the con- the contract- And I'll kind of-- Uh, sorry, Sam, I'll tack on something with that 'cause I believe in the earnings call y'all mentioned that the H-- uh, the, the Hoppers are s- are actually worth more for GPU hour than when you, uh, when, when y'all deployed them at first.

49
00:09:09.120 --> 00:09:13.670
So just on to Charlie's question, like, how do you think about that- Yeah, there are a lot of questions there. Um- [laughs] Just

50
00:09:14.660 --> 00:09:21.240
taking a, taking a step back, the contract structures have also become considerably more attractive in this business.

51
00:09:21.280 --> 00:09:30.040
We target strong returns during the initial contract term and underwrite the economics without assuming any residual value for the GPUs.

52
00:09:30.760 --> 00:09:40.460
Very importantly, customer prepayments and third-party equipment financing further, uh, reduce the amount of White Fiber balance sheet capital required.

53
00:09:41.020 --> 00:09:50.340
At, at the same time, um, evidence does continue to build that high-end GPUs can have very high useful economic life well beyond three years.

54
00:09:50.420 --> 00:10:00.080
I think-- I know that for accounting purposes they're written off after three years, but that is just false. The, the economic value of these GPUs live well beyond three years.

55
00:10:00.120 --> 00:10:07.620
But for accounting purposes, there's an ultra-conservative, and I would say unreasonable, standpoint on writing these off after three years.

56
00:10:08.120 --> 00:10:18.280
That creates the potential upside after the initial contract term, but we do not need to rely on that residual value to attract-- to generate attractive returns.

57
00:10:18.690 --> 00:10:28.880
So based on contracts signed to date, our cloud services portfolio is expected to generate more than two hundred million dollars of annualized revenue once fully deployed.

58
00:10:29.580 --> 00:10:42.900
So we have substantial contracted growth ahead of us, increasingly attractive unit economics, and a capital-efficient model for bringing these deployments online because we use customer prepayments and third-party financing.

59
00:10:43.220 --> 00:10:53.040
We don't use our balance sheet. And also, to support that growth, we entered into a partnership, um, with data center developer and operator Krembu.

60
00:10:53.740 --> 00:11:04.080
This partnership gives White Fiber Cloud access to one hundred megawatts in 2027, which is a significant expansion on our current megawatt footprint.

61
00:11:04.660 --> 00:11:17.700
So we, we have the demand, frankly, to fill all this capacity, and we're actively working to secure end place customers as we speak. So Sam, I wanna s- uh, sw- uh, shift our attention to NC1 here.

62
00:11:17.840 --> 00:11:26.800
A lot of analyst words were spilled on the call about this super quick timeline, uh, for deployment. Yeah. So y'all have got the fifty-four megawatt gross.

63
00:11:27.080 --> 00:11:37.760
Next is forty-five megawatt gross, and you're evaluating an additional two hundred megawatts for three hundred megawatt gross. Curious, the 10-Q says the Duke study supports up to two hundred megawatt gross.

64
00:11:37.820 --> 00:11:48.020
Is that just for the expanded tranche, or is that for- They-- Yeah, on top of the, on top of the second tranche that we're getting it. That's right. That would be on top of, so that's three hundred in total. Okay.

65
00:11:48.160 --> 00:12:00.360
Um, yeah, so a- uh, any updates you can give us on that study timeline here, and also, um, anything that might be standing in the way of getting that two hundred megawatts, or if y'all are pretty confident that that's gonna come through sooner than later?

66
00:12:00.980 --> 00:12:10.160
We have a great relationship with the utility company. We have a great relationship with the community, um, and that is important to have. And we have a great relationship with, with customers.

67
00:12:10.960 --> 00:12:23.876
Ju-just taking a step back, you know, this is not our first rodeo. We've been able to take Facilities that were left for dead and transform them, transform them into very sophisticated AI campuses in record time.

68
00:12:24.556 --> 00:12:33.096
And we were able to do that for Cerebras, where we took a mattress factory and we turned that into a very sophisticated tier three data center in six months and on budget.

69
00:12:33.556 --> 00:12:45.916
We were able to take a one million square foot textile factory that we bought from Unifi. We turned that into a data center for Enscale. They're about to IPO. By the way, Cerebras has since IPO'd.

70
00:12:46.596 --> 00:12:58.976
And so NC1 now has moved from construction into active customer deployment as we speak. Initial billing has commenced. Equipment startup and customer testing is underway.

71
00:12:59.656 --> 00:13:09.256
Uh, and we continue to expect the full contracted run rate billing across the full 40 megawatts of contracted IT load later this, this month.

72
00:13:10.116 --> 00:13:24.256
So, and we also, you, you, you hinted to it, uh, we see meaningful expansion potential at this campus. We expect to begin marketing the next approximate 45 megawatts of gross capacity in the very near term.

73
00:13:24.396 --> 00:13:33.036
Everybody-- we're getting very overwe- overwhelmingly strong interest for that second tranche. I know it's a very juicy tranche that everybody is after.

74
00:13:33.576 --> 00:13:47.736
Our existing customer at Enscale has certain priority rights, but we will evaluate the opportunities based on customer credit quality, based on the economics, which we think is gonna be premium, financeability, and of course, client concentration questions.

75
00:13:48.396 --> 00:14:05.936
Longer term, we are working, to answer your question directly, we are working with the utility on a transmission study that could support approximately 200 megawatts of additional power, and if completed, that will turn NC1 into approximately 300 meg- gross megawatts over time.

76
00:14:06.676 --> 00:14:13.586
NC1 remains our flagship campus. It has the potential to support several phases of growth for White Fiber.

77
00:14:14.056 --> 00:14:26.056
We intend to develop, um, that opportunity responsibly, maintain a high standard of execution, and continue to be a, a very strong long-term partner to the Madison community.

78
00:14:26.116 --> 00:14:35.566
We actually sat down with the community physically face-to-face. We had a one-day community, um, day where we engaged with the community, answered all sorts of questions.

79
00:14:35.956 --> 00:14:47.856
There were concerns and worries about our uses of water. We showed them that the way we, we manage water will be eighty-five percent less than the previous tenant. They had concerns and questions about noise.

80
00:14:47.896 --> 00:14:54.616
We showed them there will be mu- dramatically less noise than the previous tenant. That's really, you know, that's just a couple of examples.

81
00:14:54.636 --> 00:15:03.596
There are many examples of the benefits of a retrofit as opposed to greenfield, and a lot of the pushback on data centers is really against retro-- against greenfield.

82
00:15:04.056 --> 00:15:08.176
And we have a winning formula, and we have know-how in retrofits.

83
00:15:08.656 --> 00:15:21.276
And that, we think, and we've experienced, that doesn't really get into a lot of resistance with the community because that building was already there, and we can show that we could be a far better tenant than the previous tenant was.

84
00:15:21.336 --> 00:15:23.876
So it's a win-win, uh, for everyone.

85
00:15:24.276 --> 00:15:40.016
But anyway, taking a step back from the community engagement, which I think is important, and by the way, there's a CNN, um, uh, piece about it that just came out the other day about White Fiber and its unique approach to retrofits and how that aligns well with, uh, community concerns.

86
00:15:40.776 --> 00:15:44.436
Our immediate priority is just to complete this initial deployment,

87
00:15:45.536 --> 00:15:56.636
reach the contractual run rate billing, and from there, the next phase is to secure the right customer with the remaining capacity and continue scaling NC1 in a very disciplined way.

88
00:15:57.796 --> 00:16:04.916
Y- when you were talking in that response about securing the next customer, you specifically, specifically brought up, like, credit profile and financeability.

89
00:16:05.036 --> 00:16:18.916
So I do wanna ask, your, your CFO on the call, Justin Zhu, mentioned that the financing process for re-refining the bridge loan for NC1 has, quote, "taken longer than expected," and that certain contractual features would have made it easier to finance.

90
00:16:19.256 --> 00:16:28.616
Can you speak to what some of those features would have been? Yeah. Um, maybe why it's taken longer than y'all would hope to get financing here, and what's a realistic timeline for closing this financing?

91
00:16:29.176 --> 00:16:39.136
Well, we just recently got into an exclusivity with extremely well-known lenders, and we're now going through the process of diligence, and it's, it's customary diligence.

92
00:16:39.616 --> 00:17:05.915
It has taken longer than I had expected, but, um, but I'm very excited that once this is done, that'll be the turn-- the flywheel that we've been telling the markets we can do, which is find a site, get a good customer, build a thing on-- in record light speed time, and then finance it with well-known lenders and pull our equity out and recycle the capital, wash, rinse, repeat.

93
00:17:06.556 --> 00:17:18.436
If we could just show that one turn, everybody will believe that we could actually do this flywheel. So really looking forward to getting this project finance-- this permanent financing done.

94
00:17:18.976 --> 00:17:29.476
Um, I'm spending a lot of my calories and time on that one work stream, and, uh, and yeah, to-- m-more, more, please, you know, stay tuned on, on, on, uh, on that.

95
00:17:30.196 --> 00:17:42.136
But I am advised by my lawyers that whenever I talk about that particular financing, I have to say, um, that, uh, there's no guarantee that the financing will be done at all or on favorable terms.

96
00:17:42.196 --> 00:17:48.556
So that's, uh, some legal jargon I've, I've got to, uh, say. Gotta have the caveats, right? Yeah. Right. It, uh, it makes total sense.

97
00:17:48.936 --> 00:18:02.196
Sam, as a kind of a closing question, I want to hone in on the two separate business lines. Uh, y'all mentioned that the co-location and powered shell and also the cloud businesses, uh, the teams are work siloed.

98
00:18:02.216 --> 00:18:11.016
There's a potential that they might work together in the future. But White Fiber is kind of an interesting company within this universe in the sense that y'all are doing both at the same time.

99
00:18:11.356 --> 00:18:19.296
That's very rare from what we've seen from the companies that spun out from Ethereum or Bitcoin mining into this, into this universe.

100
00:18:19.856 --> 00:18:24.056
And so y-you spoke to this on the call a little bit, but I'd love for you to unpack it a little bit more here.

101
00:18:24.096 --> 00:18:30.976
And you mentioned that White Fiber blending co-location and cloud Is, quote, "An ongoing debate we are having internally," end quote.

102
00:18:31.016 --> 00:18:37.626
And I'm just curious, what does integration look like in practice, and on what timeline? Like, you've got these powered shells.

103
00:18:37.656 --> 00:18:46.896
Does that eventually mean that maybe if the economics makes sense, you move your own cloud segment into these facilities? Um, what does blending actually look like?

104
00:18:46.906 --> 00:18:56.136
And how do you balance the risk of potentially compressing whichever segment currently carries the higher multiple by balancing both of those or blending the segments in the future?

105
00:18:56.216 --> 00:19:08.376
Well, for now, like both, both-- You're right. They're both, these teams are siloed. [coughs] These teams are siloed, and i- if there's gonna be a blending of the two, it has to make economic sense.

106
00:19:08.436 --> 00:19:18.596
There's no sweetheart situation. It has to, you know, the, the, the, the cloud team has to make its case to the co-location team. And right now, we are seeing that case.

107
00:19:18.876 --> 00:19:23.716
Um, at our Montreal Two lo- location, we've restarted development there.

108
00:19:23.756 --> 00:19:37.016
We're evaluating the site for both co-location customers with potential White Fiber cloud deployments, and that does create an opportunity for us to capture economics from the data center side and the GPU infrastructure operating inside it.

109
00:19:37.036 --> 00:19:53.776
We'll be double-dipping in many ways, and we're targeting that, um, by around year-end. So that's very exciting. That'll be the first time where the teams are combining forces and making very attractive economics occur.

110
00:19:53.836 --> 00:20:07.656
And yes, I agree with you. It-- You're right. It's extremely rare for a company to have both a successful co-location business and a successful cloud business. White Fiber has both. Sam, thank you so much.

111
00:20:07.716 --> 00:20:16.876
Really appreciate you taking the time. Congrats on the quarter. Best of luck throughout the rest of the year, and we'll, we'll be checking in after Q3, see how developments are going. I look forward to that call.

112
00:20:17.496 --> 00:20:23.576
Thank you, Sam. Thanks, Sam. All right. Take care. All right. We're gonna keep on rolling.

113
00:20:23.636 --> 00:20:41.456
Next, we're gonna talk about Eiran delivering Horizon One, and then we'll have Captain Keel Ben Gagnon on for an extended interview, and then we'll cap it off with a CNN article, uh, about how data centers are helping small towns.

114
00:20:41.836 --> 00:20:55.056
But before we go to Eiran, a word from our sponsor, CleanSpark. [gentle music] We are CleanSpark, America's Bitcoin miner, a publicly traded company with the largest operating hash rate,

115
00:20:56.456 --> 00:21:08.576
powered entirely by self-operated infrastructure across four states. This is our proof of work, and we are setting the standard for what's next.

116
00:21:09.516 --> 00:21:24.436
Learn more about the intersection of energy and Bitcoin at cleanspark.com. All righty, Charlie. Long-awaited update from Eiran here.

117
00:21:24.446 --> 00:21:31.096
Headline from Blockspace: "Eiran delivers first 50 megawatt AI cloud deployment at Childress, Texas, facility."

118
00:21:32.496 --> 00:21:42.736
There was a lot of hemming and hawing, Charlie, after Eiran's quarterly update recently, that they had been behind schedule on this. From what I can tell, that's not actually true.

119
00:21:43.096 --> 00:21:50.196
I think- I think this is on schedule, right? Yes, it is. Yeah. And I don't really understand where people were coming from in saying, like, "We're still waiting for Microsoft."

120
00:21:50.276 --> 00:22:01.136
Well, the forward guidance that Eiran's given aligns with what we're seeing right now. But we'll, we'll get into the, the details there before I kind of bring the receipts for this being more on time than not.

121
00:22:02.836 --> 00:22:15.776
So specifically, this is related to Eiran's deal with Microsoft at Childress. That is a $9.7 billion five-year cloud services agreement for the compute capacity at Childress.

122
00:22:17.056 --> 00:22:34.256
The first 50 megawatts of critical IT load f- is for direct to chip liquid cooled, uh, GPUs. Separately, NVIDIA granted Eiran, quote, "exemplar," end quote, cloud status, uh, after testing the GB300s at the site.

123
00:22:35.176 --> 00:22:44.715
[lip smack] So this is just 50 megawatts. It's about 10% of the 480 megawatt gross AI capacity Eiran is targeting for 2026.

124
00:22:44.796 --> 00:22:50.876
Delivery is still pending this year for the three Microsoft phases, totaling a combined 150 megawatts.

125
00:22:51.756 --> 00:22:58.856
Roberts, quote, in the press release, points that there are, quote, "Horizons Two to Four later this year," end quote.

126
00:22:59.376 --> 00:23:11.216
So closing that 150 megawatt gap in the remaining, you know, four and a half months we have in 2026 is really a pretty big execution question on Horizon One's timing. Market loved the news.

127
00:23:11.436 --> 00:23:22.816
Shares moved between 5 to 6.5% after, um, this was announced. [lip smack] And it's-- Again, this is something that has been long awaited for Eiran.

128
00:23:22.976 --> 00:23:35.236
It's, you know, an, a cloud contract with one of the largest tech companies in the world. And Eiran's own stated guidance said that, uh, delivery was not late.

129
00:23:35.296 --> 00:23:44.036
It's within the, quote, "through 2026," end quote, window that the company has always used. Not really sure where this narrative comes from.

130
00:23:44.396 --> 00:23:54.436
In mid-June, sell side and social commentary had converged on a Q3 2026 target, roughly through July or September. We're firmly within that window on the Horizon One handoff.

131
00:23:55.076 --> 00:24:03.756
Um, apparently, there was some unconfirmed rumor that circulated, um, for a July 19th handoff date. Um, but there was nothing official behind that.

132
00:24:04.376 --> 00:24:20.026
Uh, you know, Eiran has been very explicit with this site, um, that they were expecting a Q3 2026 online date for Horizon One, so. Do you think it might have been, like, people looking at, um,

133
00:24:21.196 --> 00:24:28.105
like, satellite images and thinking that stuff's not in place on time or something like that? Maybe. 'Cause that seems to be a lot of, like, the timeline questions.

134
00:24:28.216 --> 00:24:38.556
People want to see, like, s- timestamps, satellite images, generators getting hooked up, dry-- That kind of stuff, so. Yeah, I, I wonder about that.

135
00:24:38.616 --> 00:24:42.176
You know, I, I don't have any information as to whether or not that's true.

136
00:24:42.356 --> 00:24:52.035
I will say, kind of to your point, it'd be very easy for a kind of amateur sleuth online to look at satellite images and either one, [laughs] have no idea what they're looking at.

137
00:24:52.156 --> 00:24:58.616
[laughs] Or two, not looking at the timestamp for the satellite, right? 'Cause like if you go- Yeah... to Google Earth,

138
00:24:59.616 --> 00:25:07.036
there can be a huge gap between current imagery and what is actually sh- or, uh, you know, the current state- Yeah... of things versus what's actually shown.

139
00:25:07.536 --> 00:25:19.556
There are some more expensive kind of enterprise satellite, uh, imaging services that have, like, real time updates. There's some really good ones though if, if you're interested in kind of tracking this.

140
00:25:20.336 --> 00:25:27.186
Uh, there's a company called OnX that does satellite imagery for- Man, you're blowing all of our alpha. But [laughs] keep going.

141
00:25:27.186 --> 00:25:35.716
Well, I, you know, if people are interested, they, they do satellite imagery for, uh, ex- you know, outdoor sporting enthusiasts. So, like, they have, like, an OnX Hunt

142
00:25:36.676 --> 00:25:48.156
line where you can look at, you know, Bureau of Land Management land, uh, public land for use for hunting, things like that. They also have off-road, [clicks tongue] um, and trying to think of, of the other one.

143
00:25:48.236 --> 00:26:00.796
Back Country, I think, and I think it's one of those two, On Ro- Off Road or Back Country, that has, like, pretty, pretty, uh, good, not real time updates, but they update it, like, weekly, I think, most of the satellite imagery.

144
00:26:00.856 --> 00:26:11.386
That's all you really need. I mean, yeah. Yeah. Right, I mean, like, it j- oh, you're not gonna quibble over having a 24 hour... A, a seven-day lag versus a 24-hour lag, I think, on these images. Not yet.

145
00:26:11.386 --> 00:26:15.936
But- Not until we reach peak acceleration, but yeah. [laughs] Yeah. Yeah, precisely.

146
00:26:15.996 --> 00:26:32.836
So, you know, I, I think just, g- to go back to the original thing about there being mismatched expectations here, I do think it's a reminder that you really have to dig into company disclosures if you wanna verify what some analysts and, and armchair traders are saying about these things.

147
00:26:33.516 --> 00:26:48.726
And just to recap, you know, I think the bigger question is how quickly do they move on the rest of the Horizon build-outs, 'cause they have another 150 megawatts, um [swallows], or they have another 100 megawatts to come online for, um, the, for, for Microsoft here.

148
00:26:49.556 --> 00:27:06.716
Or sorry, excuse me, 150 megawatts o- uh, um, for the rest of the year. Believe the deal's for 200. So. Yeah. I do have, uh, a very snazzy video from Iron that they dropped this morning, uh, on this announcement.

149
00:27:06.816 --> 00:27:14.836
And I don't wanna always play everybody's video, but this one's a cool video. So I'm gonna roll tape here and, uh, show it 'cause it's pretty clean.

150
00:27:15.176 --> 00:27:28.936
For those of you who are listening, I apologize, uh, it features great drone flyover oversight. [gentle music] Today, Microsoft accepted delivery of Horizon One.

151
00:27:33.176 --> 00:27:47.596
Purpose-built for AI. 50 megawatts of IT load. Direct to chip liquid cooling. Power dense design, high bandwidth fiber, engineered from day one for the densest compute clusters in the world.

152
00:27:48.616 --> 00:28:03.976
Confirmed by NVIDIA as an exemplar cloud for GB400. We developed it. We built it. Now we operate it. From engineering and procurement through construction, commissioning and handover.

153
00:28:04.496 --> 00:28:16.056
Owning the critical paths is how we deliver with speed and certainty. Thousands of people on site, every work stream managed in-house, delivered safely on an accelerated timeline.

154
00:28:16.316 --> 00:28:30.336
Horizons 2, 3 and 4 now follow through the remainder of 2026. Each build strengthens the platform. Each one shortens our time to compute. Horizon One delivered.

155
00:28:31.116 --> 00:28:44.826
[gentle music] Two, two comments here. Yeah. One, Dan Roberts giving Harry Sudock a little run for his money there as- Yeah... the in-house voiceover for company announcements.

156
00:28:45.476 --> 00:28:53.816
But of course, that's as an American, and we Americans like to fetishize- [laughs]... other Anglo accents, you know, whether it be British or Australian.

157
00:28:53.856 --> 00:29:03.036
That being said, it really is a very clean video and a rare glimpse into what some of these data centers actually look like once they're on the ground. Actually, three comments.

158
00:29:03.136 --> 00:29:09.576
Second one, gotta get the solar panels in there out in front. Yeah. You know? Like [laughs] Iron really- Go back to the roots. Go back to the roots.

159
00:29:09.676 --> 00:29:20.476
Renewable, Ir- you know, uh, Australian- Well, and that was their pitch with Bitcoin mining Yeah I mean, in the early days they were the 100% renewable energy Bitcoin miner and a lot, that [laughs] you know.

160
00:29:20.516 --> 00:29:25.316
I mean, that's how everybody, that's how, that's how CleanSpark was. Yeah, and, and- But they were more just clean energy data, you know, in general. Right.

161
00:29:25.476 --> 00:29:36.316
And I think that for, um, for Iron specifically, renewable energy credit's doing a lot of lifting there. I mean, certainly you're not powering that entire data center with solar out in Texas, right?

162
00:29:36.376 --> 00:29:49.656
I mean, great penetration of solar and wind in Texas, depending on the day, 20 to 30% of the electricity produced, but base load is king, my guy. Third thing, direct to chip liquid cooling.

163
00:29:51.496 --> 00:29:59.256
Water user, water user, uh, uh, water use FUDsters in shambles, dude. [laughs] You know?

164
00:29:59.336 --> 00:30:10.036
I mean, literally you should just send that to people who are worried about AI destroying water reservoirs because that right there just, uh, you know, throws egg onto that narrative, so.

165
00:30:11.016 --> 00:30:19.396
You say that, don't underestimate how, how deranged the Facebook groups can get, Colin. They can get very, very crazy very, very quickly. I would know.

166
00:30:19.796 --> 00:30:32.356
Um, if you're interested in following that narrative, uh, we will be posting a little bit more, uh, peek behind the curtain of the, uh, anti-data center, uh, Facebook mob on our social media.

167
00:30:32.736 --> 00:30:44.596
But we're not gonna linger on that too much because we have Captain Keel in the wings, Ben Gagnon, Keel Infrastructure, who we'll bring on up here for an interview right after a word from our sponsor, Luxur.

168
00:30:50.386 --> 00:30:56.776
This episode of Blockspace Live is brought to you by Luxur's Commander, Bitcoin miner management software for enterprise operations.

169
00:30:57.276 --> 00:31:02.866
Luxur's Commander gives you real-time fleet monitoring, bulk remote commands across your fleet, and Intelligent Miner.

170
00:31:03.286 --> 00:31:10.266
That's an automated profitability engine that runs every five minutes and adjusts your fleet's power settings against live energy and hash markets.

171
00:31:11.066 --> 00:31:16.456
ERCOT back tests show 10% improved profitability with intelligent mining versus old-fashioned binary mining.

172
00:31:17.116 --> 00:31:26.055
Commander Pro is roughly half the cost of competition, $100 per megawatt or a 25 basis point pool fee adder, but you can also try it free for 60 days.

173
00:31:26.086 --> 00:31:40.716
So if you'd like to learn more, go to luxur.tech/commander to get started. All right, we've got CEO of Keel, Ben Gagnon, in the wings, and we are gonna bring him up and unpack this Q2. Ben, welcome- Ben...

174
00:31:40.726 --> 00:31:52.246
to the show, sir. Appreciate you taking the time. It's nice to be here. Great. Always great to have you on, man. We've got a lot to chew through, so we're just gonna go ahead and dive right in. Let's do it.

175
00:31:52.746 --> 00:32:02.246
This was brought up on the Q2 call. Also, there have been a lot of questions from investors following the Q2 call, specifically about the three-deal target for end of year 2026.

176
00:32:03.026 --> 00:32:16.446
Just out in front, does that timeline still stand? Are you still confident on it? Are we stretching it out? Do you have any updates for that? Yeah. Um, look, we remain incredibly confident here.

177
00:32:16.566 --> 00:32:27.146
Um, the commercial process is going so smoothly across all three sites. Uh, we've got multiple high-quality tenants actively at three different tables.

178
00:32:27.206 --> 00:32:35.916
And everything that we've seen over the last couple of months is really only reinforcing our confidence, uh, in the assets, in the commercial process, and the demand for the power.

179
00:32:36.506 --> 00:32:43.906
'Cause, you know, when you have these conversations and, and you have a lot of different people who are reaching out to Powur, uh, it all just comes down to timelines.

180
00:32:44.126 --> 00:32:55.845
And everyone's looking at, how do you get 2027 delivery? And when you are so focused on 2027 delivery, you know, frankly, you're just a lot more time sensitive than we are as the developer, right?

181
00:32:55.905 --> 00:33:06.666
We're coming in here for 15 to, to 20 years of contracted revenue. Um, and, you know, whether or not it starts in March or April, like, really does not matter in the grand scheme of things.

182
00:33:06.746 --> 00:33:17.726
These are, you know, 15 to 20-year agreements. If the start date is plus or minus, you know, a few days, weeks, or months, it really doesn't change things in the grand scheme of things. But for customers, it's huge.

183
00:33:18.206 --> 00:33:22.166
And so each tenant we have at our table knows there are other tenants too.

184
00:33:22.626 --> 00:33:46.026
And the fact that we've been able to go through this commercial process without granting exclusivity at this stage is a huge accomplishment in itself, and I think it's a testament to the demand that we have and the power that people are willing to go through a process which is more competitive than obviously they would like because they are so eager to get the timelines that we have.

185
00:33:46.066 --> 00:33:53.966
And so, look, if you want '27 power, uh, you probably just don't have the luxury of dragging out negotiations over nice-to-haves.

186
00:33:54.026 --> 00:34:04.966
And if you do, you're either gonna push out your own timelines, or maybe you lose the megawatts entirely to someone else. And so that's what keeps real tension in the process and keeps everyone engaged.

187
00:34:05.006 --> 00:34:12.016
And if you think back to what we outlined 18 months ago, it is exactly [chuckles] the strategy that we said that we were gonna execute against.

188
00:34:12.546 --> 00:34:26.695
We said we wanted to minimize the time between signing a lease and generating the first revenues and bringing that lease online because it's going to play into the improving macroeconomic, and it's going to play into the improving power scarcity.

189
00:34:27.186 --> 00:34:38.406
It's going to play into de-risking the sites for, for Keel and our assets, and ultimately, it should drive better commercial outcomes and economics. And, uh, that's exactly where we are today.

190
00:34:38.726 --> 00:34:53.146
So yeah, we are incredibly confident on our commercial processes. So moving on into the Pennsylvania site specifically, I wanna kind of talk about permitting and, uh, environmental regs.

191
00:34:53.926 --> 00:34:57.846
A kind of multi-part question here. I'll just kinda start with th- this first off.

192
00:34:58.406 --> 00:35:09.006
I-i-if there is anything standing in the way between these leases kind of being signed and then broadcasted to the public, is it permitting and final sign-off for the sites that you have in Pennsylvania?

193
00:35:09.066 --> 00:35:11.146
Can you speak to that a bit?

194
00:35:11.206 --> 00:35:23.616
Yeah, let me kind of also just speak to, like, what is the general framework for walking through one of these commercial processes, because they're, they're highly similar across all different tenants and across all different sites.

195
00:35:23.666 --> 00:35:32.826
And the very first thing that you do is you do a quick back of the envelope check on timelines and economics, right? You're not negotiating at that point.

196
00:35:33.146 --> 00:35:42.366
You're just validating for both sides that this is something worth pursuing and is a solution for the tenants, and it is attractive for the developer.

197
00:35:43.026 --> 00:35:58.066
Once you're through that initial back check, the next thing that you do, and this is the vast, vast, vast majority of the time, the work, the conversations, it's construction and engineering and technical teams working with construction, engineering, and technical teams, right?

198
00:35:58.186 --> 00:36:12.746
So our teams work with the potential tenant teams. They work alongside the architecture and engineering firms, the general contractors, the equipment manufacturers. Everyone has to be working together in those timelines.

199
00:36:13.026 --> 00:36:27.026
And so, you know, I don't know what the exact number is, but I'd say somewhere probably between, like, 75, 85, you know, percent of the conversation is around engineering. It's around timelines. It's technical work.

200
00:36:27.366 --> 00:36:40.366
It's not negotiating of the terms. And so really the, you know, the permits remain one of the last few items that investors can noticeably, uh, see just by what's out there in the public realm.

201
00:36:40.386 --> 00:36:52.034
But it's not as if the permits are a light switch that, you know, you flip one, you get one, and then a lease is signed immediately. It's one of the last things that are remaining there for the investors to see publicly.

202
00:36:52.054 --> 00:37:07.234
But in the background, we're focusing on, on continuing with the, uh, finalizing, double-checking, triple-checking the engineering, the, uh, commercial side, the general contracting side, the equipment manufacturing side so it all lines up.

203
00:37:07.834 --> 00:37:17.254
Um, and that's, that's really all that's left in order for us to be comfortable with, with signing a lease. So, um, there's still a few more things.

204
00:37:17.314 --> 00:37:28.254
Like I said, it's not a light switch, but, um, everyone is, is very, very motivated to secure the power, and that is exactly what we're motivated to do as well, is to secure the tenants.

205
00:37:29.454 --> 00:37:39.034
So o-on that note, zooming in here, y- there are a number of [clicks tongue] uh, there are a number of permitting and environmental sign-offs that y'all need.

206
00:37:39.094 --> 00:37:47.774
You mentioned that the Panther Creek's final environmental permit on the call is, quote, "Taking a few months longer than anticipated." Why is that?

207
00:37:47.814 --> 00:38:00.994
And as a follow-up, can you give us a kind of TLDR breakdown of exactly what permitting and regulatory sign-offs y'all need for Sharon Panther Creek? Yeah. So the, um,

208
00:38:02.314 --> 00:38:10.434
like I said, you know, I think with you guys before, right, there's three general buckets of permitting. You've got zoning, you have development, and you have engineer- or environmental.

209
00:38:10.974 --> 00:38:16.854
You know, zoning tends to be your controversial one. It's the one where you have town halls, and the community engagement is very, very high.

210
00:38:17.304 --> 00:38:26.934
It's also the one where, you know, the necessary fact sets of the patter- of the project might not necessarily be what determines the outcome because it can be political, it can be emotional.

211
00:38:27.574 --> 00:38:35.454
Um, once you're past zoning, which is the controversial one, then you go through development. Development tends to be very engineering and process driven.

212
00:38:35.514 --> 00:38:45.754
It's, "Hey, you know, the setbacks require you to be 20 feet back from the property line. There are building restrictions no higher than, you know, X feet. You need to meet these kind of noise requirements.

213
00:38:45.814 --> 00:38:52.394
You need to have these kind of light requirements at different times in the day." It's not very political, right?

214
00:38:52.514 --> 00:39:04.193
It's very much a, "These are what we agreed to as kind of the frameworks and the parameters to guide construction, and are the plans that you're submitting compliant with those? If you are, great. Proceed.

215
00:39:04.654 --> 00:39:15.014
If not, these are where you're not compliant. This is what you need to revise to be, you know, um, compliant, and then you can proceed." And then the same thing with the environmental.

216
00:39:15.534 --> 00:39:20.834
Um, the environmental is very process driven. It's very data driven. It doesn't tend to be political.

217
00:39:20.874 --> 00:39:32.794
The permits that we have remaining in Pennsylvania largely have to do with stormwater waste runoff and connecting into municipal sewer systems. These are not politically controversial permits, right?

218
00:39:32.834 --> 00:39:37.094
We're not talking about flushing millions of gallons of water a day into a municipal sewer system.

219
00:39:37.624 --> 00:39:46.714
We're talking about making sure the employees on site have the normal facilities that any facility would have, and they're connected into the municipal sewer system.

220
00:39:46.724 --> 00:39:57.414
Um, we're talking about making sure that the standard plans for any kind of development, um, are there for stormwater waste runoff and how that's gonna be controlled and managed with the, with the layout of the land.

221
00:39:57.534 --> 00:40:08.784
So these are pretty standard. They're pretty straightforward, and there's nothing to be read into on the delays. There's just two different departments in Pennsylvania for the Department of Environmental Protection.

222
00:40:09.274 --> 00:40:18.094
There's an eastern office, which serves obviously the eastern side of the state where Philadelphia and the vast majority of the population and the industry is.

223
00:40:18.134 --> 00:40:25.883
And then there's a western office which s- you know, serves a much smaller population and a much smaller industrial base, which is where Sharon and Scribd Grass are.

224
00:40:26.414 --> 00:40:33.794
And so at Sha- uh, the western side of the state, fewer applications, fewer permits to work through, and they have less of a backlog.

225
00:40:33.834 --> 00:40:40.274
On the eastern side of the state, way more permits to work through, a bit less, a bit more understaffed, and they have a bit more of a backlog.

226
00:40:40.694 --> 00:40:49.014
But we haven't received anything in the permitting process that would give us, you know, pause, concern, alarm, and nothing for the tenants as well.

227
00:40:49.114 --> 00:41:00.524
Um, these just continue to remain moving forward at the same time that we're running the commercial process, and we have not had one concern from the tenants around the remaining permits at either of those sites.

228
00:41:02.194 --> 00:41:17.314
So I'll throw a question in here. Um, looking at Moses Lake in Washington, uh, I think it's your smallest site, 18 megawatts, and it looks like you, uh, did not exercise the option to expand it- Yes...

229
00:41:17.354 --> 00:41:26.454
another 10 megawatts, which would, would have been 28. What's the reasoning for that? I'm, I'm kinda curious. Everyone wants more power. Why are we not doing more power?

230
00:41:27.594 --> 00:41:34.374
Uh, so it's a great question, and let me walk through kind of the reason why we got the option in the first place and why we walked away from it.

231
00:41:35.094 --> 00:41:40.434
So, you know, obviously Moses Lake at 18 megawatts is the smallest site that we have under active development.

232
00:41:41.054 --> 00:41:47.694
Um, and it just, because of the scale, limits itself down to, you know, a different kind of, of potential customer.

233
00:41:48.354 --> 00:42:04.614
What we, you know, thought would be a very attractive, you know, um, kind of, you know, sweetener to any sort of a, um, a, a tenant contract in Moses Lake would not only be to have 18 megawatts but would be to have the ability to expand beyond that and continuing growing.

234
00:42:05.034 --> 00:42:16.614
Because at every site that we've seen with every tenant that we've spoken to, they don't just want first power, but they want a relationship to grow over time with and to scale over, over a longer period of time.

235
00:42:17.134 --> 00:42:24.094
And so we thought it would be an attractive thing for us to do, acquire a low-cost option, have that optionality in the commercial process.

236
00:42:24.574 --> 00:42:33.194
If tenants would be interested in the site but maybe they need a little bit more power, we would have that there to provide greater optionality and greater value to the tenant.

237
00:42:33.934 --> 00:42:44.562
Um- But what it's come down to is, is that like I said earlier, it's all just about timelines, and the 10-megawatt option was not for a, a 10 megawatts adjacent to our site.

238
00:42:44.582 --> 00:42:51.132
It was a 10 megawatts at a different site in the same town. Um, it was going to be associated with much longer timelines.

239
00:42:51.642 --> 00:42:58.382
And one of the things is, is, with smaller sites is a lot of the costs just don't amortize as well over way fewer megawatts.

240
00:42:58.962 --> 00:43:06.711
Some of the things are fixed regardless of if it's 18 megawatts, 100 megawatts, or 350 megawatts, it's the same amount of work.

241
00:43:07.171 --> 00:43:14.251
And so, you know, what we would rather do is we, you know, we brought that to have optionality for the tenants and provide extra value for the tenants.

242
00:43:14.892 --> 00:43:28.542
But we would rather focus on the better opportunities that we have in our portfolio to acquire and secure larger sites, which don't have the same kind of economy of scale disadvantages of something like a 10 megawatts.

243
00:43:28.582 --> 00:43:44.042
We'd rather focus on bigger sites, um, especially in the East Coast, Northeast, and the Midwest area, where we think inference is going to be, uh, the highest, you know, demand market, and that's where we wanna try and, and focus our efforts moving forward.

244
00:43:44.122 --> 00:43:52.342
So, um, we think it's, you know, the right thing to do to continue to grow the pipeline and allocate resources internally, uh, effectively.

245
00:43:53.862 --> 00:44:07.202
Uh, going back to Pennsylvania, you talked a bit about permitting, um, and I want to ask a question about the Sharon site, which if it looks like you've collapsed the 30 and 80 megawatt phases into one

246
00:44:08.162 --> 00:44:19.062
single 110 megawatt phase, um, what's behind that decision? Uh, yeah. Yeah. So, uh, that's what we're working towards right now.

247
00:44:19.182 --> 00:44:24.442
We're, we're pretty close to thinking that we're, we're gonna have that finalized and done, but it's not a, a done deal today.

248
00:44:25.082 --> 00:44:32.262
Um, the reason why is because even the proximity between the racks themselves have huge latency impacts.

249
00:44:32.702 --> 00:44:41.342
And so, you know, the closer that we can get the compute to itself, the better it is for the performance of the site and for the tenant.

250
00:44:41.352 --> 00:44:52.722
And so what we wanted to do was we wanted to try and consolidate the power density as much as possible. The original design was a little bit more, uh, spread out across, you know, multiple buildings.

251
00:44:53.082 --> 00:44:59.591
Here, we're trying to consolidate it in closer and closer so that it is, you know, an ultimately better product for the end tenants.

252
00:45:00.062 --> 00:45:06.962
Um, we think that we've accomplished that with, uh, our engineering, or we're about to accomplish that with, with our final engineering plans.

253
00:45:07.082 --> 00:45:15.282
Um, and we look f-forward to talking about that publicly once it's final, you know, with the new updated plans, the site images and the renders and everything else.

254
00:45:15.862 --> 00:45:26.622
Uh, but ultimately it was just trying to make a better product for the customers. So it's kinda like the Jensen scale up, now you're trying to scale out- Correct... uh, idea. Gotcha. So Ben,

255
00:45:27.662 --> 00:45:34.172
with regards to leases in, uh, uh, you know, potential leases in the pipeline, you, you've said before that

256
00:45:35.282 --> 00:45:49.642
companies like Keel are kind of in this Goldilocks zone, and I kind of read that as you're not a first mover, but you've gotten to see how certain deals have been moved across the finish line, and we're actually seeing dollar per megawatt deals for powered shells move upwards.

257
00:45:50.202 --> 00:45:58.022
I mean, the, the, the infamous example of this is C- Core Scientific with the CoreWeave deal. Great deal at the time. CoreWeave, you know, paying for the build-out.

258
00:45:58.382 --> 00:46:05.542
But if you actually look at the unit economics, there have been better b- deals struck since, and you can say the same about some of the earlier deals with Terawulf and Cipher as well.

259
00:46:06.042 --> 00:46:19.382
We're starting to see, uh, you know, this segment kind of feel out what makes the most sense, and it seems like they're able to juice more from these deals as the timeline goes on and we continue to move into this CapEx cycle.

260
00:46:20.182 --> 00:46:33.762
Long-winded way of asking this, and I know you can't give too many specifics, but with regards to a structured deal, what kind of economics and deal structure for these leases is Keel privileging?

261
00:46:33.802 --> 00:46:40.182
Like, what would you like to see in terms of terms, duration, et cetera? If you could just, like, give us a teaser.

262
00:46:40.222 --> 00:46:47.762
I know you can't speak to specific numbers, but just curious where y'all's head's at in terms of how you structure these things. Yeah. You're-- It's-- I mean, it's a great question.

263
00:46:47.802 --> 00:46:56.841
You're right, I can't speak to specific numbers. But, you know, we, we talked about this 18 months ago. The economics were gonna continue to improve. They have, right?

264
00:46:56.962 --> 00:47:12.942
Um, and it's not a perfectly straight line, but if you chart out every single lease that has been announced, you know, and you put it on kind of an apples to apples footing, like a per megawatt basis or per megawatt per month basis, wherever you wanna look at it, like, you'll see that the direction is pretty abundantly clear, right?

265
00:47:12.982 --> 00:47:21.362
Like, the economics have continued to improve up and to the right, uh, for the last 18 months. And so we do think that there is a, a second mover advantage.

266
00:47:21.902 --> 00:47:27.542
Um, you know, we also can avoid a lot of the mistakes that, you know, people have learned the hard way, right?

267
00:47:28.082 --> 00:47:49.142
Um, you know, we saw earlier on what, what happened with people giving exclusivity too early on in the commercial process, and ultimately it resulted in massive, massive delays because you grant someone six months exclusivity and you shut yourself off from everyone else, then you have to grant the next person six month exclusivity starting from zero, and then the next one.

268
00:47:49.162 --> 00:47:56.022
And so, you know, we've learned that, um, from our peers, you know, the good and the bad of what's happened.

269
00:47:56.062 --> 00:48:05.582
We've seen the market evolve, and we've been able to, you know, kind of play into what we've seen as the trend for economics and continued power scarcity to our advantage.

270
00:48:06.122 --> 00:48:18.922
So, um, you know, th-that-- with that overall framework in mind, um, what I can speak to in terms of deal structuring is, you know, we prefer triple nets. We're not trying to, you know, uh, be

271
00:48:19.882 --> 00:48:32.822
crazy in creating creative new structures. You know, we're not looking to sell, you know, rum raisin or pistachio cream. Like, we like vanilla. Vanilla is totally good. Um, vanilla sells all over the world.

272
00:48:33.222 --> 00:48:38.262
It's popular everywhere. Everyone likes it. It's easy to understand. You don't need to get too exotic.

273
00:48:38.612 --> 00:48:50.038
Sometimes simple is better, and I think for the investors, they prefer something that's easy to model, easy to understand, easy to forecast, minimal variables. So that means triple net.

274
00:48:50.698 --> 00:48:53.178
We like triple net structures across the board.

275
00:48:53.818 --> 00:49:14.638
Um, and we should expect to be getting, I think, triple net structures across the board, with one exception being Moses Lake, and that's just because the customer profile there is so different than at Sharon or at Panther Creek, and they really need a more turnkey solution, where the larger players are looking to take a lot more, um, you know, operational involvement and control.

276
00:49:14.648 --> 00:49:26.418
And so, you know, we're able to, uh, proceed on triple net, which is the preferred option for 99% of our portfolio. For the 1% of our portfolio that's Moses Lake, we'll take the structure which is best for the...

277
00:49:26.458 --> 00:49:39.498
you know, maximizing value for that particular site. Um, but overall, we are very focused on triple nets. I'm, I'm gonna start demanding more food analogies from other executives. Also, my wife in shambles.

278
00:49:39.528 --> 00:49:45.958
She thinks that vanilla is a boring flavor. I think vanilla- She's wrong. I actually like vanilla more than most other- Vanilla rules... fruit. Yeah. Vanilla freaking rules.

279
00:49:46.218 --> 00:49:52.388
Everyone likes vanilla except for your wife, I guess. So- [laughs]... it's generally a, a, a pretty good way to go, I think.

280
00:49:53.338 --> 00:49:59.338
So Ben, I have one more question for y'all specifically, and then I wanna zoom out to the sector at writ large.

281
00:49:59.878 --> 00:50:12.578
Curious about lead times for equipment and whether or not Keel has moved forward for securing long lead time items, whether or not you're seeing any snags within the supply chain considering there's so much demand.

282
00:50:12.958 --> 00:50:18.138
You could just speak to that, uh, the supply chain logistics side of things very quickly. Yeah, I can.

283
00:50:18.178 --> 00:50:27.658
And, you know, slightly different, um, at each site, but j- by and large, what we've done is, uh, Moses Lake, absolutely everything is secured. It's contracted.

284
00:50:27.678 --> 00:50:34.098
It's either received on site or it's being manufactured somewhere else, and we're waiting to receive it on site in the coming months.

285
00:50:34.218 --> 00:50:43.238
Um, with regards to Sharon and Panther Creek, what we've been doing, this is what I spoke to earlier in the commercial process, you run all of that simultaneously.

286
00:50:43.278 --> 00:50:50.978
Like, you have to run two parallel paths because different customers could have different equipment choices, and they could wildly differ.

287
00:50:51.118 --> 00:50:55.838
Um, there are big changes, especially when it comes to things like backup generation solutions.

288
00:50:55.878 --> 00:51:06.478
That's usually one of the big areas where people have very differing opinions and equipment choices, and that can represent, you know, $2 to $3 million a megawatt in terms of CapEx.

289
00:51:06.498 --> 00:51:10.268
So that's a big one that you don't want to put on your balance sheet unnecessarily.

290
00:51:10.838 --> 00:51:26.957
And so what we do is we go parallel paths with the tenants, with the general contractors, the NE firms, and most importantly, those equipment manufacturers to make sure those supply chains and the, the timelines are all lined up, and then you ideally execute all at the same time.

291
00:51:27.378 --> 00:51:45.608
What you don't wanna have is you don't wanna have, you know, uh, an unmatched liability where we've just signed up to procure, you know, $3 billion of equipment that we f- you know, don't have the tenant for, and we also don't want to sign up for a tenant to deliver $3 billion of equipment that we don't have secured.

292
00:51:46.178 --> 00:51:58.638
And so the way that you try and balance this is by concurrently running those processes in parallel and then getting to the point where you can sign all the agreements all at the same time and not have those unmatched liabilities.

293
00:51:59.118 --> 00:52:14.738
And at the stage where we are right now, working through with our Go Vertical partners and the equipment supply chain, we're hitting those earliest RFS dates in 2027, um, with a little bit of ramp into 2028 for Sharon and then a ramp through 2028 for Panther Creek.

294
00:52:15.238 --> 00:52:31.438
So the timelines that our customers want, first power online in '27 and continuing to ramp from there, uh, we're very confident we can achieve those timelines, um, based on all of the work that we've done to date, all of the contract negotiations we've done, and all of the work with the supply chain manufacturers.

295
00:52:31.718 --> 00:52:41.758
But that's, that's generally the framework how you want to, uh, to run that. I appreciate that. All right. Last question from me. I'll toss it to Charlie for a closer. I know we gotta get you out of here soon.

296
00:52:41.798 --> 00:52:51.268
So you said on the call that you're, quote, uh, you're watching, quote, "The increasing amounts of intercompany financing across the industry," end quote, as a top risk.

297
00:52:51.978 --> 00:53:04.427
What specifically concerns you, and how does this affect Keel's approach to tenant credit support? And also, what financing options y'all would entertain once you have a lease on the books? Yeah.

298
00:53:04.438 --> 00:53:18.347
So I think that came up in, um, analyst Q&A in response to a question of, like, what is your biggest risk factor for, for Keel rolling forward? And, um, you know, the answer that we gave was, we think it's macro, right?

299
00:53:18.448 --> 00:53:30.038
The stuff that the company is, is executing against, we are incredibly confident in. We know that the demand is there for the power. We know that we are making the progress that we want on the permits.

300
00:53:30.618 --> 00:53:43.678
So we are very confident in what we are in control of. I think the, the one thing that, you know, is constantly changing in this space is the macro environment, right? Um, and it's changing on a, on a daily basis.

301
00:53:44.418 --> 00:53:53.438
Monday is something new. Tuesday is completely different. Wednesday is somehow different than that. And by the time you get to Friday, it feels like a month has passed, and it's been five days.

302
00:53:53.538 --> 00:54:05.838
So, you know, those are the things that we really, you know, try to, try to keep a closer eye on, um, is just what is going on in the broader financing market. This is a very capitally intensive industry.

303
00:54:05.878 --> 00:54:21.438
The cost of financing these projects can really change the ultimate economics of the project themselves, and we wanna make sure that we're hitting the market at the times that we wanna hit the market, where we can get the best economics and the best deals.

304
00:54:21.458 --> 00:54:33.658
And sometimes, you know, high-yield market will just shut down. Sometimes it's really, really cheap to do a convertible. Sometimes there's a tremendous amount of low-cost debt. Sometimes there's not.

305
00:54:33.938 --> 00:54:42.478
And so what we are keeping an eye on, you know, is this macro factor of what's going on in the world. How is this impacting the broader industry?

306
00:54:42.518 --> 00:54:55.350
What does that have in terms of implications for, uh, supply and demand side of the equation? And, you know- We are really focusing on how do we get the most value out of what we have in our pipeline.

307
00:54:55.370 --> 00:55:09.130
And so understanding that macro piece, understanding the timing component of these things, we think is really mission critical. So to wrap up, kind of a broad industry question about AI models.

308
00:55:09.450 --> 00:55:22.830
Um, open models have been having a bit of renaissance. Jensen had that big letter, NVIDIA, and pretty much everybody signed except for a conspicuous couple companies. Um, in a world where open weight models,

309
00:55:23.730 --> 00:55:36.130
uh, become much more prevalent, maybe even dominant, what happens to the market? Who wins in this scenario? Who loses in this scenario? Yeah, what does this look like in your view?

310
00:55:36.530 --> 00:55:40.690
Um, you know, anyone's guess at this point is, is really as, as good as mine.

311
00:55:40.790 --> 00:55:52.150
Um, but I think what we're seeing is we're seeing there is starting to be a lot of, uh, segmentation in the market, and people are starting to become more and more niche players, right?

312
00:55:52.330 --> 00:56:03.120
You're seeing platforms come out, um, like Higgsfield, which is trying to establish itself as like the premier like image video generation platform, you know.

313
00:56:03.170 --> 00:56:12.170
And then you've got that as completely separate from Claude for Enterprise, which is completely separate from ChatGPT, which is more consumer friendly or consumer directed.

314
00:56:12.690 --> 00:56:18.910
And then you've got, you know, the open weight models out of China. How all of this is gonna evolve, I don't think anyone really knows.

315
00:56:18.930 --> 00:56:49.780
But I think what it shows is that there's gonna be more than one model dominating across the market, because as with everything in this world, uh, people like to specialize, people like to focus in, and people generally don't want to have, you know, like, um, one solution for absolutely everything in the box, because you can't be the best solution for everything in the box when you're a solution for absolutely everything, right?

316
00:56:49.830 --> 00:56:53.890
If you are a company, you don't have, you know, one system, right?

317
00:56:53.930 --> 00:57:03.890
You may have payroll, you may have a legal system, you may have a business development system, you have a customer relationship management system, you've got finance systems, you've got payroll systems, right?

318
00:57:03.990 --> 00:57:14.870
It's not, I think, reasonable to think that one system is going to rule all of those systems. I think we're gonna see increasing fragmentation, specialization of these models start to emerge.

319
00:57:15.450 --> 00:57:32.810
Um, and I think it's a great example of Jevons paradox playing out here, because the reality is, is that, you know, there's always different ways that you can do price segmentation, and you can try and capture more value for premium services or lower value for lower quality services, right?

320
00:57:32.850 --> 00:57:45.270
This is the-- you know, this is a very proven economic model. You know, airplanes sell, uh, economy, premium economy, business, first class. Trains do a similar thing. Cruise ships do this.

321
00:57:45.690 --> 00:57:58.130
You know, the China model has been, um, focusing on low cost volume, you know, in terms of just how they produce goods and services, and the Chinese as, or the US system has always been focused on higher quality, higher value.

322
00:57:58.650 --> 00:58:05.440
So in some situations, maybe you're like an, you know, a defense industry contractor. You prob- maybe don't want the open weight model.

323
00:58:05.910 --> 00:58:14.950
Maybe you want the highly specialized model for, for you, which is going to be maybe it's Claude, or maybe it's something you've developed specifically for yourself.

324
00:58:15.070 --> 00:58:26.370
Um, but I think the more applications that you see, the more models that you see, the more optionality that there is, people will start to do price segmentation, um, more and more.

325
00:58:26.950 --> 00:58:37.890
And I think, you know, you're gonna see enterprises focus more and more on the higher quality models, and you're gonna see retail gravitate towards the cheaper Chinese models.

326
00:58:37.950 --> 00:58:43.730
Because if, you know, for us as a business, um, I don't want slop, right?

327
00:58:43.810 --> 00:58:54.970
[chuckles] Like we, we use Claude in our management team, and, you know, we, uh, encourage everyone to be asking Claude questions and trying to, to, to play around with Claude as much as possible.

328
00:58:55.530 --> 00:59:00.429
We don't want anyone running anything through, you know, uh, open weight models right now.

329
00:59:00.490 --> 00:59:09.390
Like, we want everything on an enterprise system that is, you know, secured, that is safe, that is, you know, signed off by our IT team, is reliable.

330
00:59:09.869 --> 00:59:18.890
Um, but you know, if I go home and I wanna generate a, you know, cute picture of a dog or something, you know, I'm not gonna be using Claude, right? I'm gonna be using a different model.

331
00:59:19.030 --> 00:59:31.270
I think it's just showing how early we are and the, the future is gonna be increasing fragmentation, um, and specialization. Ben Gagnon, CEO of Keel, Captain Keel himself. Thank you for joining, man.

332
00:59:31.440 --> 00:59:42.570
Really appreciate it. Best of luck with the rest of the year, and we'll have to run it back Q3 when y'all have results then as well. Thanks, guys. Looking forward to it. Thanks for your time, Ben. Thanks, Ben.

333
00:59:43.170 --> 00:59:55.670
One quick thing here. A lot-- I mean, the troll box is popping off on this one today. Yeah. We see the comments. We love the comments. I, I want to just highlight this from Joke in a Box, or Jock in a Box, great name.

334
00:59:56.240 --> 01:00:04.270
"Guys, help the channel out. If this gives you value, like it, like it, et cetera." It helps them out, like leaving a good server a tip. That's not one of us in a- That's not one of us.

335
01:00:04.450 --> 01:00:13.370
We really appreciate- That's, I promise that's not a spoof. Yeah. Yeah, we really appreciate that. Um, thank you, Jock in a Box. Subscribe to this channel, give us a like, comment, um, a- anywhere you find it.

336
01:00:13.450 --> 01:00:22.190
Share it with people who are also invested in these buckets of assets. Uh, this is the end of our earnings week's coverage almost. We've got one more interview tomorrow.

337
01:00:22.200 --> 01:00:30.230
We also have Hive Digital's results, uh, coming in at the end of market, uh, market close tomorrow, and then there's a live stream that we're gonna be doing on Monday.

338
01:00:30.270 --> 01:00:36.190
That being said, we've had interviews with a number of CEOs. We've live streamed most of the major earnings calls.

339
01:00:36.230 --> 01:00:42.570
You can go to our channel, go to live, and you can find that, and you can also find clips from the CEO interviews if you're interested. Yeah.

340
01:00:42.730 --> 01:00:58.930
There's a playlist with exec- called Executive Interviews on the YouTube, where we, uh, aggregate all the executive interviews we've had, not just during earnings week, but just in general 'cause we do c- we have these interviews all throughout the year, not just earnings week, so Thanks for tuning in I'm 100% on that.

341
01:00:59.230 --> 01:01:00.330
Yeah, thank, thank you, guys.

342
01:01:00.430 --> 01:01:12.170
And one last thing I'll say with regards to Keel, so Ben mentioned in the earnings call Q&A section with analysts that the management team has a betting pool on which site will get a lease first.

343
01:01:12.690 --> 01:01:21.220
Based on that interview, if I were a betting man and I were at Keel, I'd probably put it at Moses Lake, but not financially, like- Well, it sounds like he- I mean, he sounded pretty confident.

344
01:01:21.250 --> 01:01:29.559
He's like, he's pretty- Oh, yeah... like, he's like, "We've no... It's a particular, it's a particular, you know, client profile," so that sounds like it's this one.

345
01:01:29.559 --> 01:01:34.389
Yeah, Moses Lake definitely seems closest to initially- But it's the smallest. It's the pi- it's the one people don't care about. Yeah.

346
01:01:34.430 --> 01:01:38.350
And it's the one- They're all, they're all like, "Three deals of the year," which I get, guys, but, you know.

347
01:01:38.890 --> 01:01:43.670
Yeah, and I know there are a l- a, a lot of folks who are kind of frustrated with not having a clear timeline on these things.

348
01:01:43.850 --> 01:01:52.850
Uh, my read from this though is that if they can get this permitting done this year, I think we're going to get leases announced sooner than most people think.

349
01:01:52.930 --> 01:02:02.300
Ben's going to be conservative, because he doesn't want to give a specific date and then be in hot water later, even if he was talking about getting three across the finish line earlier in the year.

350
01:02:02.370 --> 01:02:10.550
I understand people are frustrated, but, um, anyway, I think that does it for Keel, Charlie. Yeah. Uh, stick around.

351
01:02:10.590 --> 01:02:23.390
We're gonna do a lighthearted story about why data centers are not the worst thing in the entire world, and in fact, actually can bring good to a community. Uh, but before that, a word from our sponsor, Ligos.

352
01:02:23.490 --> 01:02:34.190
[static] 64K, gang, is haunting your dreams.

353
01:02:34.730 --> 01:02:43.250
If you've invested in Bitcoin over the last year, no doubt you're a little disappointed by its performance, especially con- uh, when you, if you weigh it against AI stocks.

354
01:02:43.270 --> 01:02:53.770
But you'd be even more disappointed if you used a Bitcoin-backed lender that ended up fumbling your bag because they were rehypothecating your funds. There's a long track record in the history of this happening.

355
01:02:53.890 --> 01:03:03.570
Do yourself a favor. If you are looking for a Bitcoin-backed lender or you currently use one, you need to check out Ligos Finance. They're our preferred Bitcoin-backed lender here at Blockspace.

356
01:03:03.590 --> 01:03:09.510
They use Bitcoin native smart contracts to make sure you are always in control of your keys. That's right. You heard that correctly.

357
01:03:09.530 --> 01:03:15.370
With Lig- Finance, you never relinquish control of your keys while you are pledging collateral for a Bitcoin-backed loan.

358
01:03:16.050 --> 01:03:29.550
With Ligos Finance, there's no rehypothecation, no wrapping, no bridging, no funny business, just competitive rates as low as 10.5, 8.5 to 10% APR. Go to ligos.finance to learn more. All right.

359
01:03:30.090 --> 01:03:49.710
We're gonna get dragged if this video gets to, uh, this clip gets to your neighborhood Facebook group, but for now, uh, we're gonna close with a CNN article, uh, out this morning profiling how data centers transformed the economy of this small town.

360
01:03:49.790 --> 01:03:55.770
The small town in question, Quincy, Washington, a rural farming town

361
01:03:57.210 --> 01:04:10.430
of roughly 8,500 people in Central Washington, which has seen major public infrastructure upgrades funded largely by data centers, one of which is a $15 million aquatic center,

362
01:04:11.850 --> 01:04:29.430
an upcoming Q-Plex indoor sports complex, and a, am I correct reading this column, $120 million high school? Yeah. State-of-the-art high school valued at 120 million US dollars.

363
01:04:29.570 --> 01:04:40.619
If, I struggle to see how a high school... Anyway, um. I mean, there, there's probably a consideration there where maybe they're, the, the county is financing that, and then they plan to pay down the financing- Yeah...

364
01:04:40.670 --> 01:04:45.030
with tax revenues from- Yeah... the site. Right. Yeah. So let me keep going. Where is this money coming from?

365
01:04:45.090 --> 01:04:55.360
It's coming from roughly 30 data centers, um, that started with a Microsoft data center back in the 2000s, later joined by Yahoo, Dell, Intuit, and others.

366
01:04:55.890 --> 01:05:09.490
These data centers pay an estimated, sit down, 57% of the city's property taxes. They, why are they there? Because of cheap, abundant hydroelectric power from the Columbia River.

367
01:05:09.670 --> 01:05:23.390
So over half of the property taxes, I should have scoped out where Washington and this county sits in terms of property taxes, but I can't, like in relative to the field, but I can't imagine it's, uh, the cheapest in the world.

368
01:05:24.170 --> 01:05:29.440
Um, yeah. In Washington, absolutely not. That's why I'm saying. It's Washington, king of... Yeah. Yeah. We're gonna tax some good notes.

369
01:05:29.440 --> 01:05:35.250
As someone who lives just south of Washington in the Communist Republic of Oregon, I can tell you [laughs]

370
01:05:35.990 --> 01:05:42.630
that depending on where it is, now if you're in the more rural areas, the property taxes actually tend to be s- uh, lower than the national average.

371
01:05:42.670 --> 01:05:53.030
Well, it doesn't, if you're in, if you're in, uh, Quincy, Washington, it probably doesn't matter as much, 'cause mo- because your property, the property taxes are being paid by the data centers, and they're going to community development.

372
01:05:53.060 --> 01:06:06.050
And what's, what are some other upsides? 900 direct jobs, engineers, pipe fitters, four to six indirect jobs each. Uh, uh, yeah. Um, so what are...

373
01:06:06.110 --> 01:06:17.590
A- another crazy stat is the poverty rate in this county dropped from 29.4% in 2012 to 6.2% in 2024. What? Like, you can, you can, um...

374
01:06:18.650 --> 01:06:35.370
When, when poverty rates are going up, uh, in Western countries, particularly the United States, um, dropping poverty rate, not just incrementally, but substantially, uh, is massive. There's new housing, new businesses.

375
01:06:36.170 --> 01:06:51.910
Um, the local officials call it the Quincy Miracle that turned this low-opportunity farming town into a brand-new everything town. Um, what's the downsides? I love a brand, a brand-new everything town. Yeah.

376
01:06:51.950 --> 01:07:07.512
What are the downsides of this? Um, well, uh, power rates rose, but they only rose by 3.5%. Um, industrial parts- Also, uh, just a quick, a c- a quick caveat there. I, I really wonder about this specifically when we...

377
01:07:07.542 --> 01:07:19.302
So this is, i- in Oregon, power rates are going up here as well. Uh, Oregon is one of the, in terms of concentration of data centers, one of the largest in the US, at least before this AI boom.

378
01:07:19.362 --> 01:07:32.502
Part of this, though, is that, number one, both of these states have started prioritizing solar panels and wind over base load, and if you're in the southern part of Oregon, that's fine. It's sunny here all the time.

379
01:07:32.981 --> 01:07:42.182
It, we get a lot of sunshine. Uh, it's not as rainy as it is in Portland and Seattle. That being said, they're also starting to decommission dams, hydro dams in the area.

380
01:07:42.602 --> 01:07:50.862
Some of that is for environmental reasons because they're worried about salmon. Uh, f- some of it is just Looney Tunes, as far as I can tell.

381
01:07:51.262 --> 01:07:56.902
So when you see something like that, m- there's m- there might be some truth to the fact that the data centers are driving it.

382
01:07:57.002 --> 01:08:08.862
I will say, though, a lot of the policies that have been put in place over the last decade have been explicitly, like, anti-growth on the energy front, right? Yeah. Kind of- Well, they're trying-...

383
01:08:08.872 --> 01:08:20.342
in these utopic- Yeah... green, green energy fantasies. Yeah. Well, they're trying to basically ensure that there's no more Quincy miracles happen. God forbid, uh, your town would get a little- [laughs]... property tax.

384
01:08:20.402 --> 01:08:25.822
God f- God forbid my to- [laughs] God forbid a man get a little data center tax- Yeah... revenue. Okay.

385
01:08:26.062 --> 01:08:34.182
So, so I've been, I've been kind of dismissive, uh, 'cause the, the, you know, uh, basically data centers bring tons of money, jobs, everything. Um,

386
01:08:35.582 --> 01:08:48.682
however, if you've lived in the town for forever, this is a story pretty much any place in the world which ex- which experiences economic boom town, this happens everywhere. Um, rising housing costs.

387
01:08:49.222 --> 01:08:59.562
You might call this gentrification. It's not necessarily that. It's just demand in this case, not people coming in and making everything way too nice so you have to move out.

388
01:08:59.602 --> 01:09:10.532
But housing costs are up, uh, pretty significantly. Housing costs have roughly doubled since then. This, th- sorry, can I piggyback on this for a second, Charlie? Yeah.

389
01:09:10.562 --> 01:09:21.962
To me, this is the clearest negative for data center builds for local communities. Yeah. Um, the electricity price thing really isn't that big of a, uh, isn't that big of an issue.

390
01:09:22.042 --> 01:09:36.452
If you look nationally, at least, on average, when a data center moves into an area, actually they create grid stability because the utility now has a bulk buyer of energy, and they can potentially fire up additional assets, some that have been remaining idle.

391
01:09:36.922 --> 01:09:45.372
It actually helps with the economics more often than not. The water use thing is just a total non-starter. It, it is a complete straw man for what's actually happening. It's a ridiculous point. Yeah.

392
01:09:45.382 --> 01:09:53.202
I don't know why people keep making it. Uh, just their brains being- And, but no one's making the housing argument, and that's a- Yeah... that's a huge one because number, there are a few things at play here.

393
01:09:53.722 --> 01:10:05.602
These towns are often very small. I mean, Quincy, the latest census was, like, just south of 8,000. So these are areas that have really no industrial base. You're bringing in these data centers.

394
01:10:05.642 --> 01:10:12.461
There's suddenly a huge surge in demand for housing, specifically when they're being built. So you kinda have this,

395
01:10:13.522 --> 01:10:29.482
this interesting equation to where they become boom towns when they're being built, and that rent is going to get driven up, but that rent's probably going to cascade back down at a certain point because the, the times at which these data centers employ the most people is when they're being constructed.

396
01:10:29.542 --> 01:10:38.582
I mean, you saw the Iron video that we played earlier in the segment, uh, earlier in the show. Thousands of people. We said thousands of people. In the middle- Yeah... of nowhere, Texas. I mean, yeah. I- I- if you go...

397
01:10:38.652 --> 01:11:07.062
So back to the Facebook, uh, the Facebook Boomer Facebook group, um, they hate what they call, uh, like, man camps, where it's, like, a bunch of Winnebagos out in the middle of, you know, nowhere of a bunch of, uh, dirty, blue collar, salt of the earth working guys out there, you know, uh, laying copper and, uh, and welding and, uh- And grilling steaks after- Oh, terrible...

398
01:11:07.092 --> 01:11:12.902
a long day of construction. Yeah. But, but- Cracking Busch lights. But, like, think about it, though. I mean, this is, like, a, a huge issue.

399
01:11:12.942 --> 01:11:19.922
You got a bunch of, uh, probably single, uh, transitory guys who are there for four to six months.

400
01:11:20.202 --> 01:11:34.422
Granted, they're working all the time, but I, it, it might be fun to be the owner of the local watering hole for a couple months, but then, man, it... If y- just imagine. And, and, you know, it, I, I totally get, um,

401
01:11:35.882 --> 01:11:38.462
uh, the housing thing. Uh, I don't know how you solve that.

402
01:11:38.702 --> 01:11:48.452
That is kind of a- I, I don't think you can solve it other than, like, maybe some towns try to strike a deal with the developer saying, like, "You need to have temp housing." Some- I, I don't know how you do that.

403
01:11:48.562 --> 01:11:58.312
That would be add to the burden here. I, I think the simplest ac- uh, to your point, the simplest thing that they can do probably is just have, like, these kind of,

404
01:11:59.302 --> 01:12:08.482
you know, like you said, these itinerant camps of workers coming in, and maybe you just have them there. You know? It's, like, basically a big music festival, just a bunch of RVs and big [laughs] tents. Yeah.

405
01:12:08.502 --> 01:12:14.742
It, it might have to, it might have something to do with perhaps considering the full cycle impact of the, of these sites.

406
01:12:15.042 --> 01:12:23.232
Um, less think about it more narrowly from just property taxes today, but how do you offset rising costs of rent?

407
01:12:23.282 --> 01:12:33.712
'Cause if you consider that you have people in these towns who have been living there for, for a long time who are renting, like, that, it, that's just a terrible situation to be in. I hate to see that. Yeah.

408
01:12:33.712 --> 01:12:41.522
For those who own homes, like, it might, they might enjoy their house going up. If their property taxes go up, uh, they're not gonna- They're not gonna enjoy that...

409
01:12:41.532 --> 01:12:47.352
they're not gonna enjoy- Especially when you consider the median income in these towns is often at or below the national average- Yeah...

410
01:12:47.402 --> 01:12:56.382
because they're just re- there aren't in- there is not really much industry in these towns. There's, there's, you know, there's not really any job growth, especially in the Pacific Northwest.

411
01:12:56.462 --> 01:13:06.358
I hate to keep hammering this, but, like, if you look at national employment statistics, like, Oregon and Washington are- Really struggling with regards to adding their- Why?

412
01:13:06.378 --> 01:13:14.098
They're tech- they're tech capitals of the world... not Oregon. What's going on there? Seattle kind of is. Oh, Seattle. But like, the fact of the matter is there are a lot of social problems.

413
01:13:14.238 --> 01:13:23.818
I mean, no one wants to build in cities where quite frankly like there's so much social disruption, and you have homeless people everywhere openly using drugs on the sidewalks, you know.

414
01:13:23.878 --> 01:13:28.758
So, um, not- We've turned this into a different kind of podcast [laughs] Yeah, not, not to hammer on that too much.

415
01:13:28.898 --> 01:13:39.618
I wanna highlight one other instance of data center prosperity though here, and we kinda covered it on the show I think a week back. But specifically Applied Digital in Ellendale, uh, North Dakota.

416
01:13:40.538 --> 01:13:50.748
Ellendale is a very tiny city, and their pre-data center tax revenue, um, or sales tax take was roughly $400,000 a year.

417
01:13:50.778 --> 01:14:04.478
In the first seven months of 2026, the town collected 3.5 million from this data center that Applied Digital has erected in the area. That's 9X its old full year total.

418
01:14:04.918 --> 01:14:15.718
That was used for street repaving, a senior center renovation, new public safety complex, walking trails, and a restoration of the town's historic opera house. You know,

419
01:14:16.738 --> 01:14:26.558
I, I do just think that it is important to highlight these, because I think that, number one, there is so much negative press regarding these data centers, particularly within the legacy media.

420
01:14:27.378 --> 01:14:32.648
The fact of the matter is the reality is going to be somewhere in the middle. There are negative externalities.

421
01:14:32.738 --> 01:14:41.298
Rising rents in some of these large town- of these small towns can, uh, significantly impact the local tax base. That should be listened to.

422
01:14:41.678 --> 01:14:49.398
But the fact of the matter is you have towns that are, some of them are dying and floundering. No one wants to live there. People who grew up there move out.

423
01:14:49.438 --> 01:15:00.698
Now they actually have money to improve the town, and I, I ca- it's hard for me to say that that's not a huge positive if weighed against some of the negative benefits. I also think that it's an interesting

424
01:15:01.878 --> 01:15:16.178
example of the fact that if you look back in the '50s and '60s, the wealthiest town, the wealthiest cities in the US were in the Rust Belt in terms of per capita income and per capita net worth.

425
01:15:16.698 --> 01:15:23.848
And then Tulsa, Oklahoma 'cause of oil money, but yeah. Yeah, okay. Yeah, yeah. Okay, good, good example. I mean, that, you could also say this with oil and gas boom as well. Yeah.

426
01:15:23.858 --> 01:15:30.338
But if you look at that, like things like Detroit and Cleveland, they, they had some of the highest standards of living and were super wealthy.

427
01:15:31.758 --> 01:15:48.038
I, I think that you're going to have a similar scenario probably over the next few decades with regards to some of these data centers, at least in terms of the tax base, that, that the, the states and the towns that privilege these data centers are going to reap fruits that are obvious in hindsight-

428
01:15:48.978 --> 01:15:57.418
Yeah... compared to the places that are shunning them. Uh, comment in chat, "Once locals figure out how much money is available for the community, they'll beg for the data center to build there."

429
01:15:57.498 --> 01:16:06.158
I think we'll see many such cases. I think it's gonna be all over the place as, uh, it's probably deal specific. We cover this a lot.

430
01:16:06.258 --> 01:16:28.338
Like, a lot of these are gonna be deal specific, and, um, I think if the people can wake up to the reality that, um, you might be able to use your, you know, your zoning and your permitting, uh, bottlenecks as s- really as, as leverage against these folks, not to kick them out or prevent them from building, but rather to say, "We want you here.

431
01:16:28.698 --> 01:16:37.068
You've got to pay, uh, your diligent- your, your dues to the town." Because they have money. And as- Yeah...

432
01:16:37.068 --> 01:16:43.678
as permits become like the constraining factor, if it requires them to just commit more cash towards public infrastructure,

433
01:16:44.738 --> 01:16:54.458
s- uh, certain, uh, you know, certain initiatives, uh, I think you'll see people who are very, very happy. Um- Yeah, it's probably- Um... it's very deal specific. Yeah.

434
01:16:54.658 --> 01:17:06.048
Uh, one last comment from me and then we can, uh, put this horse down. Uh, there's a lot to talk about here obviously. Yeah. But I, I do... So that comment came from YouTube, Go- Goog. Hell yeah.

435
01:17:06.048 --> 01:17:10.898
And there's some fantastic names in the troll box today. I agree with that in part.

436
01:17:11.358 --> 01:17:27.348
I will say, I think we have reached such a polarized point in our political discourse that I think that people actually privilege moral arguments over substantive arguments with regards to what's actually beneficial.

437
01:17:27.478 --> 01:17:30.148
I wouldn't be surprised if you saw some really obdurate

438
01:17:32.738 --> 01:17:45.078
citizens and obdurate, uh, you know, political factions that irrespective of the benefits that these data centers could bring, they're just gonna oppose them on vague moral grounds that might not even stand up to scrutiny.

439
01:17:45.798 --> 01:17:46.138
But

440
01:17:47.258 --> 01:18:03.178
s- some people will just straight up not want them, uh, for, for, uh, for the same reason why certain people, uh, don't like to do, d- don't wanna clean up their cities or institute law and order, because they think that it's, you know, kind of propping up something that they don't believe in.

441
01:18:03.238 --> 01:18:04.598
Anyway- I think it's funny...

442
01:18:04.638 --> 01:18:20.558
I'm, I'm kind of b- Carl lives in a commie state and is more conservative than me I live in a commie state, but I- I live in a, in a, in a r- in a very, very red state and I'm more commie than Ca- and I think we're the, we're the communist ones See, I live in a commie state, but in a locality that is very libertarian, don't tread on me.

443
01:18:20.618 --> 01:18:30.238
So I, I... It's, it's just, uh, because- Which is funny 'cause I live in, I live in the most purple or blue county- You have-... in the reddest state [laughs]

444
01:18:30.298 --> 01:18:38.137
Well, it's interesting to see the way that the, uh- We're probab- we're probably like, uh, right, perfectly ideally, like ideologically aligned because of this [laughs] Yeah.

445
01:18:38.158 --> 01:18:47.558
There's like, we're like, our, our, our areas are like diametrically- Yeah... uh, it's, it's like the inverse of what the other one is, is seeing. Yeah. But you do get to see the push and pull of both- Yeah...

446
01:18:47.638 --> 01:19:00.058
arguments as a result. So anyway. Yeah. Okay, enough about that. Thank you for listening to Block Space Live. We do this live stream every single weekday at 1:00 PM Eastern. We are Compute's daily live show.

447
01:19:00.098 --> 01:19:08.838
So if you're interested in compute, data centers, AI, emerging technology or markets, tune in every single weekday, 1:00 PM Eastern.

448
01:19:10.198 --> 01:19:26.198
If you can't get enough of the live stream, it's a podcast wherever podcasts are found, and if you can't get enough of that, you can go to our website blockspace.media where we do regular daily coverage of everything, investor notes, bank notes, new news, deals signed, yada, yada, yada.

449
01:19:27.398 --> 01:19:37.538
This show is brought to you by CleanSpark, NASDAQ listed, ticker CLSK. And we have one more day of earnings week coverage.

450
01:19:37.578 --> 01:19:52.538
Tomorrow we have Mike Piacchi on from Soluna, as well as the Soluna earnings live stream later today on our channels. Make sure to not miss that. See you tomorrow. I'm Charlie. I'm Colin. And we'll catch you tomorrow.

451
01:19:52.718 --> 01:20:01.378
[outro music]