OM DENNE EPISODE
We track why markets struggle even as AI fundamentals keep getting stronger, from Bitcoin’s move to a surge in Treasury yields that changes the financing math for data center builders. We also connect the dots between power shortages, rising GPU pricing, and blowout Micron results that validate just how fast the AI infrastructure cycle is scaling.
• Bitcoin strength and what it signals for proxies
• 10-year Treasury yield spike and how it hits consumer spending
• Higher yields raising the cost of debt for AI infrastructure companies
• Why contracted backlog and blue-chip counterparties matter most
• FTC probe, AI safety pledges, and the speed of model releases
• The scale of AI capex compared with prior infrastructure revolutions
• Corvex as an emerging compute and security-focused data center story
• Power demand shortfalls and the long-lease path toward REIT-like trading
• GPU price inflation and why CSP and co-location economics improve
• Micron earnings, HBM expansion, and what it implies for NVIDIA demand
• Street price targets for MU, CoreWeave, and NVIDIA
Smash a like button, guys. If you're not already subscribed, McNally Money, feel free to join. Let us know in the comment section below how you're feeling about the portfolio and if you're making any purchases at these levels!
Anthonys Info:
Website:
https://www.powermininganalysis.com
X:
https://twitter.com/cazenove_uk
Patreon:
https://www.patreon.com/Cazenove_UK
Bryce's Info:
X:
Patreon:
https://www.patreon.com/McNallieMoney
Merch:
https://shop.mcnalliemoney.com/
Website:
Business Inquiries:
Email - powermininganalysis@gmail.com
VIS NOTER 🔗
UDSKRIFT 🔗
00:00:00.000 --> 00:00:04.639
Hey guys, welcome or welcome back to the channel McNally Money, the official home of power analysis.
00:00:04.879 --> 00:00:11.519
Market's struggling once again today, despite some big fundamental catalysts that we're gonna talk about in the presentation.
00:00:11.759 --> 00:00:12.640
A lot to go through.
00:00:12.800 --> 00:00:15.679
Before we do, take a second, smash a like button, guys.
00:00:15.759 --> 00:00:17.199
Big help to myself and the channel.
00:00:17.359 --> 00:00:18.719
Anthony absolutely loves it.
00:00:18.879 --> 00:00:22.160
If you're not already subscribed, McNally Money, feel free to join.
00:00:22.239 --> 00:00:28.480
And let us know in the comment section below how you're feeling about the portfolio and if you're making any purchases at these levels.
00:00:28.640 --> 00:00:31.519
With that being said, let's get into today's video.
00:00:40.320 --> 00:00:41.840
All right, guys, away we go.
00:00:42.000 --> 00:00:47.759
A big episode in store for you today, Thursday afternoon, the first day of October, Anthony.
00:00:47.840 --> 00:00:52.079
Hopefully you've got the Halloween costume picked out and a big day in the markets.
00:00:52.159 --> 00:01:05.040
We've got a number of headlines, micron earnings, which were absolutely phenomenal, markets recovering, and a number of big catalysts in terms of demand, GPU pricing, and uh price targets.
00:01:05.120 --> 00:01:06.560
So this should be a good one, Anthony.
00:01:06.640 --> 00:01:13.280
We're gonna start off with Bitcoin and then some industry news, then we'll move into the miners and AI infrastructure players.
00:01:13.359 --> 00:01:16.959
We've got Bitcoin on the one day here, up over 84,000.
00:01:17.120 --> 00:01:24.239
And we were just saying before we started, a lot of the DATs or Bitcoin proxies having a pretty good start to the day.
00:01:24.640 --> 00:01:30.719
Yeah, um, Bitcoin's actually, you know, uh rallying close to 85,000 on a live perspective.
00:01:30.879 --> 00:01:36.480
So um, but uh yeah, having having a better day and getting to the close to the week's highs there.
00:01:36.640 --> 00:01:40.319
So um, yeah, positive, positive green day for Bitcoin.
00:01:40.640 --> 00:01:49.200
And we've been saying, I think this was the seventh day in a row the AI infrastructure stocks were looking to be red, almost overdue for a green day.
00:01:49.280 --> 00:01:53.359
We know tomorrow's option expiration, so hoping for a big rebound there.
00:01:53.519 --> 00:02:05.760
We'll get into that in a second, but one uh headline that caught our attention this morning, Anthony, the 10-year treasury yield making some big moves, now getting back near 2002 levels.
00:02:06.079 --> 00:02:07.120
Yeah, absolutely.
00:02:07.280 --> 00:02:10.879
The uh the biggest monthly jump since 2022.
00:02:11.520 --> 00:02:27.280
Um but when you think uh go back to the 2002 highs, 5.31%, the benchmark yield rose over 50 basis points just in September, and that's at the largest increase um obviously since that September 2022 uh level itself.
00:02:27.520 --> 00:02:37.120
Now, per fund strikes, hard ECA seeing valuations historically start compressing past that threshold, forcing investors, corporations, consumers to redo the mass and everything.
00:02:37.199 --> 00:02:55.919
So you'll see uh you know spending reduced now because um these levels here will impact uh spending in terms of you know, people with mortgages will will see uh you know potentially bigger rises or loans bigger rises in their payments that means less available cash to spend on the nice things.
00:02:56.080 --> 00:03:02.080
So, whenever this sort of thing happens, there think about all those discretionary costs that you you spend on a monthly basis.
00:03:02.240 --> 00:03:06.240
Are you able to continue with those discretionary costs?
00:03:06.479 --> 00:03:12.800
Some of the nice things to have, or is it a case of I've got to get the rent, I've got to get food, uh, I've got to get to work.
00:03:13.120 --> 00:03:15.439
Do I have a coffee every day at work?
00:03:15.520 --> 00:03:18.319
Um, or do I take a coffee with me to work in a flask?
00:03:18.400 --> 00:03:19.199
Blah blah blah.
00:03:19.280 --> 00:03:28.800
You know, you've got to sort of like work out that then, and we'll see a pullback now on um on spending to uh to sort of like improve the situation.
00:03:28.879 --> 00:03:55.840
Now, treasures are posted a medium loss of 0.9% in September and 0.7% in October over the past decade, um, as the bond market is already under pressure, and the market's now price uh a sort of 35% odds of an October Fed hike down from 50% earlier this week, even as elevated oil prices are still keeping inflation expectations elevated.
00:03:55.919 --> 00:04:02.479
Remember, the oil price effectively over well over a hundred dollars a barrel and really impacting everything.
00:04:02.560 --> 00:04:16.879
If you think about you know, not just uh you know, uh fuel for the for your vehicle, uh oil impacts literally everything that you purchase in terms of uh whether it's part of the actual product or part of getting that product to you.
00:04:17.040 --> 00:04:26.000
Um so when the oil price is high, uh that impacts inflation, probably bigger than any other um part of the basket that they use to calculate inflation.
00:04:26.319 --> 00:04:34.720
Well, and we were talking, Anthony also impacts savings because uh once you're starting to get these levels of yields, maybe people put money into treasury bonds.
00:04:34.879 --> 00:04:39.839
So we've got interest rates fluttering around, we've got treasury yields.
00:04:40.079 --> 00:04:45.680
The question for you is how does this impact the companies we care about or cover on a day-to-day basis?
00:04:45.839 --> 00:04:49.920
These companies are going out there, they're raising tremendous amounts of money.
00:04:50.160 --> 00:04:52.240
We know how interest rates impact.
00:04:52.399 --> 00:04:55.120
How does this yield impact the companies as well?
00:04:55.360 --> 00:04:57.519
Yeah, it's gonna impact them in a number of ways.
00:04:57.680 --> 00:05:06.319
These higher yields, it's gonna, you know, the price of your debt on every convert or revolver um um uh interest product there.
00:05:06.560 --> 00:05:12.240
Remember, core we recently took out convertible note 2.875%, Wi-Fi at 5%.
00:05:12.800 --> 00:05:16.240
Then you've got the compressed multiples on long duration cash flows.
00:05:16.399 --> 00:05:24.240
That's gonna hit the likes of Riot, Terra Wolf, Core Scientific Keel, harder than revenue generating, iron and bit deer.
00:05:24.639 --> 00:05:33.360
Also, we've got tougher convertible note economics and a 5.31 risk-free rate competing directly for that speculative capital.
00:05:33.680 --> 00:05:52.720
Now, the offset um, you know, the contracted backlog, uh, you've got companies like Hutt, Wolf, and and Riot sitting on tens of billions of dollars in sign multi-year revenues with the likes of Anthropic, Google, Microsoft, which is why the financing hasn't dried up despite this big move in the yield.
00:05:52.879 --> 00:05:58.399
Um, but you know, if we talk about the bottom line, there's a real headwind sector-wide ahead of us here.
00:05:58.560 --> 00:06:15.120
Uh, the names with the contracted um investor grade back revenue like HUT, uh, TerraWolf, and BitDear, will absorb it better than pre-revenue AI pivots, those that have yet to sign a contract in terms of uh like Keel, Mara, and Saluna Holdings.
00:06:15.199 --> 00:06:22.319
Now we know Saluna have got um an LOI in place there, um, and they suggest it's with a NeoCloud.
00:06:22.480 --> 00:06:27.600
So hopefully we'll see in the not too distant future that convert into a contract.
00:06:27.680 --> 00:06:48.959
But quite rightly, if you haven't got a contract at the moment, um this sort of headwind will create an issue because you've got to go out there and borrow once you've signed the contracts, and um, you know, you're going to be hit with some of these increases in potential uh products to drive that capital for your company to deliver these uh sort of these facilities for clients.
00:06:49.279 --> 00:06:51.120
Yeah, very interesting dynamic.
00:06:51.199 --> 00:06:52.879
I guess the good news is there.
00:06:52.959 --> 00:06:59.360
Uh hopefully we don't get a rate increase in October here, in light of what the Treasury Yield is doing.
00:06:59.519 --> 00:07:02.079
Now, again, big picture here, Anthony.
00:07:02.160 --> 00:07:06.879
We talked about the White House tech executive meal or lunch earlier this week.
00:07:07.040 --> 00:07:13.920
We've got an image here, pretty much an all-star cast uh in terms of who's who in the tech finance world.
00:07:14.160 --> 00:07:17.199
Interesting timing for the next update.
00:07:18.000 --> 00:07:34.560
Yeah, um, you know, this was like only a couple of days ago, and then yesterday we've got effectively the uh the FTC have now opened uh a probe into anthropic open AI and other AI labs over potential rogue agents.
00:07:35.040 --> 00:07:51.920
Now, um, you know, the FTC plans to provide subpoena like requests uh compelling you know executives to uh provide testimony um within the coming weeks, examining the possible unfair or deceptive practices under the FTC Act.
00:07:52.079 --> 00:07:55.279
I don't think this is looking out there, looking to out there for violations.
00:07:55.439 --> 00:08:02.560
This is really um enforcing what the president said is that we already have departments that are going to be running with this.
00:08:02.800 --> 00:08:15.519
The president also wants to create um this sort of like this group of about 10 special individuals that will oversee um and drive this safety issue um around AI.
00:08:15.759 --> 00:08:41.679
Now, in terms of the OpenAI agents uh you know uh recently probed and attacked the AI hub around about July 11th to the 13th, which we covered on the channel back then, um, OpenAI not realizing its own agents were responsible until after the breach was actually disclosed, and it was found that 700 of the 1200 supposed isolated agents had actually joined um after coordinated on an unsanctioned message board.
00:08:41.919 --> 00:08:50.559
Um, so they they've joined the group together having having um you know just uh uh communicated with each other and then started this breach there.
00:08:50.720 --> 00:09:07.200
Now uh you know Chairman Ferguson says, you know, agents must follow instructions and and do not act with wills of their own, meaning the responsibility stays with the developers, and he's accused uh the labs of stoking fear to shape the favorable regulation.
00:09:07.679 --> 00:09:21.039
And like I say, this has all happened 24 hours after anthropic, open AI, Google, and many of the other biggest companies in the world have signed the president's voluntary moral-binding AI safety pledge.
00:09:21.200 --> 00:09:24.720
Fergus himself attended that very, very same lunch.
00:09:24.960 --> 00:09:31.120
Yeah, interesting timing or ironic, but I guess Trump making a statement there that we have these agencies in place.
00:09:31.279 --> 00:09:39.039
We're gonna come on to the scale of the AI industry, but I just wanted to pause you guys that how quickly the adoption is happening here.
00:09:39.200 --> 00:09:42.720
You and I were looking at some military applications before the show.
00:09:42.879 --> 00:09:45.919
We're seeing the governments launch their own large language models.
00:09:46.080 --> 00:09:47.919
Obviously, these frontier labs.
00:09:48.080 --> 00:09:52.720
I really think we're on the cusp of something a lot bigger than most people realize here.
00:09:52.799 --> 00:09:57.840
Now, despite that FTC probe, it doesn't appear it's really slowing down.
00:09:58.000 --> 00:10:07.200
A lot of the companies developing these models, we've got another model launched today, this one from Google, right after OpenAI's dev day.
00:10:07.360 --> 00:10:09.360
We talked about Astra 6.1.
00:10:09.519 --> 00:10:12.000
These models are coming fast and quick.
00:10:12.559 --> 00:10:13.360
You're quite right.
00:10:13.519 --> 00:10:21.840
Gemini 4 Argonne and this comes in the same week that you you know the uh the Astra uh model was launched there.
00:10:22.159 --> 00:10:26.320
Also, uh Meta announced news for small businesses.
00:10:26.480 --> 00:10:34.080
Um, OpenAI also revealed uh dots, and Robin Hood uh released a number of agentic features for training as well.
00:10:34.320 --> 00:10:48.879
So, you know, AI uh releases are coming fast and furious, and we were having a conversation before the podcast price about how much out there there is, and how do people choose which one is the right one for them when there's so many products out there?
00:10:49.120 --> 00:10:55.519
I mean, we all like competition, but um, you know, this comes at a cost, and you want to make sure that you're getting the right one.
00:10:55.600 --> 00:11:08.320
And from what I'm reading, uh some of the people that we follow uh their stories regularly are sort of like not just using one of these products, they're using products to do different parts to help them with their businesses.
00:11:08.399 --> 00:11:14.559
So they might use three or four different AI um packages to develop certain aspects of their businesses.
00:11:14.720 --> 00:11:29.679
So um it's a real uh you know, uh a compet competitive marketplace at the moment, and and the fact that Google have just launched um this Argan uh AI model here is is obviously just adding to that.
00:11:29.840 --> 00:11:36.399
Now, in terms of um this particular product, it sets a new mark in the real-world software engineering.
00:11:36.480 --> 00:11:44.559
Uh, it ties for first in cybersecurity and leads a separate benchmark spanning finance, legal, and professional tasks.
00:11:44.720 --> 00:11:55.279
Um, Argon's freed up hundreds of terabytes of data center memory internally without actually producing additional hardware and also being used by quantum computing for researchers.
00:11:55.440 --> 00:11:58.879
So it's optimizing Google's own infrastructure.
00:11:59.039 --> 00:12:09.519
Now it's launching first to trusted cybersecurity partners while Google works with the US government on a pre-release safety evaluations before a broader public release.
00:12:09.600 --> 00:12:30.320
So you're going to get a phase rollout starting with these big cyber defenders, and it matches OpenAI's GPT 6 Astra, which was released this week, and GROC 7 4.7 on cybersecurity benchmarks and leaves both uh GPT 6 Astra and recent anthropic models on the Valve's index.
00:12:30.480 --> 00:12:37.200
And again, this has come out uh you know 24 hours after they all met uh in the White House.
00:12:37.519 --> 00:12:43.039
Yeah, we're gonna talk about Corvex a little later, but as you're talking there, we had the co-CEO on last night.
00:12:43.120 --> 00:12:50.399
He was saying the same thing internally, they use some of the cheaper models for day-to-day tasks, some of the more high priority tasks.
00:12:50.480 --> 00:12:53.200
Maybe they ship those out to anthropic or open AI.
00:12:53.279 --> 00:12:54.720
So more on that in a second.
00:12:54.879 --> 00:13:01.519
Uh, this slide here, Anthony, really paints the picture of what I just hit on, how big the AI revolution is.
00:13:01.679 --> 00:13:11.360
Uh, you can see, even compared to railroads, electrification, highways, this is a much bigger investment, a much bigger uh change to the world.
00:13:11.440 --> 00:13:21.679
And if you think back, Anthony, how pivotal railroads were, how pivotal electricity is, and now the fact that they're saying AI is going to be even bigger, uh, it's pretty crazy.
00:13:22.000 --> 00:13:24.480
Yeah, this appears to be the biggest revolution.
00:13:24.559 --> 00:13:36.720
Going back through history, we've had the railroads we've had the canals, uh, electric when that was um invented there, highways, and then you know, one of the biggest in recent times is recent uh telecom and fiber.
00:13:36.879 --> 00:13:37.840
Think about the internet.
00:13:38.320 --> 00:13:58.960
Um, well, AI um over this next sort of seven, eight years is planned to be at least three times bigger than the internet was when that was uh created back you know over 20 years ago, and um no stopping there, massively bigger than railroads and every other um effectively uh infrastructure.
00:13:59.120 --> 00:14:06.000
When you look at the amount of annual uh spending as a as a percentage of GDP, that's how they've calculated this.
00:14:06.159 --> 00:14:07.919
Um, so significant spending.
00:14:08.000 --> 00:14:13.919
We've been talking about you know the demand for this, but we've also talked about the cost to deliver it.
00:14:14.080 --> 00:14:54.960
You know, we we some of these companies that we talk about on a daily basis, um, you know, are spending 12 to 16 million dollars a megawatt just on building the buildings that will support the GPUs, and then the GPUs themselves are costing 30 to 40 uh million dollars each, and um and every you know few months we're seeing a new more efficient GPU come to the market, um, and you know the price increases with that there, and as we're seeing prices increase, we'll see GPU prices increase because you can bet that the chip manufacturers will want a share of that future profit as well.
00:14:55.039 --> 00:15:06.159
So the likes of NVIDIA must be loving the fact that rates are increasing with great regularity because it means they can put their prices at a level where they're making effectively super profits.
00:15:06.240 --> 00:15:18.159
And we covered NVIDIA um you know in the last few months when they brought out their unbelievable earnings and will come on to another company in the space with their earnings results from yesterday, equally as impressive.
00:15:18.399 --> 00:15:30.799
Yeah, and you think 50-60 million dollars a megawatt for CSP, that's why we care so much about rate uh percentages, rate hikes, the yield, it all impacts the financing side of this.
00:15:30.960 --> 00:15:33.759
Now we're gonna move over to the stocks here.
00:15:33.919 --> 00:15:38.960
I gotta say, Anthony, we took this screenshot about an hour ago and it looks quite different now.
00:15:39.120 --> 00:15:40.559
We've seen a pretty good bounce.
00:15:40.720 --> 00:15:43.519
I mentioned we've had a number of red days in a row.
00:15:43.600 --> 00:15:45.440
Maybe we're overdue for a green.
00:15:45.519 --> 00:15:51.919
Uh, so you can walk us through the heat map stocks at this point, but keep in mind you guys looking a little different now.
00:15:52.320 --> 00:15:55.279
Yeah, it looked like it was going to be a predominant red day again there.
00:15:55.360 --> 00:15:59.120
You can see from the day range, stocks have pulled back from their highs.
00:15:59.279 --> 00:16:05.919
Um, no uh, you know, other than what we've talked about in the podcasts already, you know, um maybe the impact there.
00:16:06.000 --> 00:16:16.080
But remember, fundamentally, these stocks are sat in a really strong position at the moment because they've got power, they'll all get contracts, there's no doubt about that.
00:16:16.159 --> 00:16:26.080
If you're thinking about 33 megawatts and then 78 to 82 megawatts demand by 2029, think about the power that these companies have got.
00:16:26.399 --> 00:16:29.200
That doesn't even eat into that amount of demand there.
00:16:29.279 --> 00:16:31.759
So they'll all get contracts at some point.
00:16:31.840 --> 00:16:36.000
It's a case of will they be able to execute on them contracts and deliver it.
00:16:36.159 --> 00:16:46.960
And from what we're seeing already, in terms of the big players who have been energizing and have been handing over power to their uh to their customers.
00:16:47.120 --> 00:16:56.559
Uh, recently we saw Iron hand over Horizon One to Microsoft, a hyperscaler, uh, with a further three buildings to be handed over by the end of the year.
00:16:56.639 --> 00:16:58.960
That's Horizon 1 to 4, fully handed over.
00:16:59.200 --> 00:17:01.360
Horizon 506 are currently being built.
00:17:01.440 --> 00:17:05.359
That'll be the next tranche of contracts being awarded for that particular area there.
00:17:05.440 --> 00:17:08.640
They've got 300 megawatts at sweet water already energized.
00:17:08.880 --> 00:17:14.480
Buildings are already started to uh take place on that particular site there.
00:17:14.640 --> 00:17:26.319
So uh the other one I'd like to mention as well is Core Scientific, probably the company with the largest amount of compute power contracts at the moment, over one gigawatt of contracts in place.
00:17:26.640 --> 00:17:30.720
And that first contract with CoreWeave at 590 megawatts.
00:17:30.799 --> 00:17:33.599
Well, they've pretty much handed over three courses of that already.
00:17:33.680 --> 00:17:38.640
The remaining part will be handed over in the next few months or very, very early into 2027.
00:17:38.960 --> 00:17:54.240
Um, but you can see um how what advantages that gives to Core Scientific, and we don't need to know too much about the quality there because we'll cover uh Core Weave um you know separately later in the podcast and give you an indication of what the market feels about core weave.
00:17:54.480 --> 00:18:04.319
Remember, core scientific are the are the landlord of of literally most of their power, so uh that should be a good indication of how well that company's doing as well there.
00:18:04.400 --> 00:18:16.880
But if you look at the uh the day range there, again, one company um it seems to be booking the trend, um, and this is a pure um uh data center company, AI density and they haven't got any background in Bitcoin mining.
00:18:17.039 --> 00:18:19.599
Uh Corvex Incorporated, you were quite right.
00:18:19.680 --> 00:18:38.960
We we had this co-CEO on uh Seth uh last night, and um, to be honest with you, for a first podcast last night, um, a very intriguing background um and story to this company and to Seth with his experience over 20 years of experience of working with data centers.
00:18:39.119 --> 00:18:46.160
His resume reads like Microsoft, AOL, Google, um, oh, and and and NASA.
00:18:46.400 --> 00:18:56.640
So we can actually really we can put um we can put Seth in the sort of um uh rocket scientist bracket uh and and and mean that because he's actually got that much experience.
00:18:56.799 --> 00:19:03.119
In terms of patents, he's had 180 patents awarded for work he's done there.
00:19:03.359 --> 00:19:04.880
180.
00:19:05.200 --> 00:19:07.599
Um, he knows what he's doing in this sector.
00:19:07.680 --> 00:19:09.200
It was an intriguing podcast last night.
00:19:09.359 --> 00:19:13.440
If you haven't watched the podcast and you don't know anything about Corvex, just have a have a listen to that.
00:19:13.599 --> 00:19:21.359
Have a listen to about the products that they're bringing out to the market in terms of security, safety uh for their customers as well.
00:19:21.519 --> 00:19:26.400
That's their uh one of their sort of like uh differentiators, if you'd like to say.
00:19:26.480 --> 00:19:31.119
Um, but it's well worth it's you know, probably half an hour, 40 minutes podcast last night.
00:19:31.279 --> 00:19:33.759
Take the time to listen to it because it's a new company.
00:19:33.920 --> 00:19:42.880
In terms of power, they're growing to about eight megawatts of um compute CSP contracted power by the end of the year.
00:19:43.200 --> 00:19:45.920
Don't think of eight megawatts as something small.
00:19:46.079 --> 00:19:56.640
I only have to go back four and a half years ago to when Iron uh approached me to start covering them when they had half an X Ash of Bitcoin mining, and and they've been in my data ever since.
00:19:56.880 --> 00:20:11.680
Iron and now a 16 billion plus dollar company, uh, you know, having some of the biggest uh clients in terms of Microsoft as part of the as part of their bailiwick, and they've grown and grown and grown in a very short space of time.
00:20:11.759 --> 00:20:15.440
Who's the next sort of diamond out there that we that we're looking for?
00:20:15.599 --> 00:20:21.440
Maybe it could be uh Corvex uh Incorporated with a ticker symbol um M O V E move.
00:20:21.599 --> 00:20:28.319
Um they've been sort of top of our um of our daily trading range ever since all the stocks started to go down.
00:20:28.400 --> 00:20:38.480
It really hasn't impacted them too much, maybe because they only uh IPO'd earlier this year, so they've not been around uh too long for people to take note of.
00:20:38.640 --> 00:21:06.640
Um, but uh with a market capitalization of$371 million, and it means that they have got uh certain levers that they can pull to grow, and they have got expansion plans not just for this year but for next year too, to significantly grow that compute CSP model, uh, very similar to what Iron are doing with Microsoft, with um a security feature uh that comes as part of the package.
00:21:06.880 --> 00:21:10.559
Yeah, and yesterday we talked about the launch of the token factory as well.
00:21:10.640 --> 00:21:17.839
So rather than renting GPUs by the hour, you can consume tokens individually, uh more tokenization kind of model.
00:21:17.920 --> 00:21:19.680
But this is an interesting one, Anthony.
00:21:19.759 --> 00:21:28.319
And we heard earlier from IRIN they're contracting megawatts now 25, 30, 40 million dollars on short-term contracts.
00:21:28.480 --> 00:21:31.839
Eight megawatts adds up pretty quick in terms of revenue.
00:21:31.920 --> 00:21:34.160
Uh now you just threw out some big numbers there.
00:21:34.240 --> 00:21:42.400
You mentioned 33 gigawatt shortfall in 2028, anywhere from 72 to 82 gigawatt shortfall by 2029.
00:21:42.640 --> 00:21:44.880
This was from an interview on CNBC.
00:21:44.960 --> 00:21:46.240
Now we've got a clip for you.
00:21:46.319 --> 00:21:50.880
They're talking about TerraWolf, a number of interesting points covered here.
00:21:51.039 --> 00:21:56.880
Specifically, pay attention to the duration of the leases and the trading over the summer months.
00:21:57.039 --> 00:21:59.279
We'll play you the clip, then we'll get your thoughts.
00:22:00.160 --> 00:22:12.079
So you said that um our estimate of 26 to 20 night 2028 incremental power demand rises to 97 gigawatts in our base case, and the shortfall is 57 gigawatts.
00:22:12.160 --> 00:22:16.799
Now to put that in perspective, because a lot of people are watching Back to the Future, and that's if the car went back in time.
00:22:17.039 --> 00:22:20.319
New York City uses between seven and ten gigawatts a day.
00:22:20.480 --> 00:22:20.720
Correct.
00:22:20.960 --> 00:22:26.480
So we are five to eight New York Cities short of power.
00:22:26.720 --> 00:22:29.759
But yet the power generation stocks until today have been going down.
00:22:30.000 --> 00:22:30.400
How come?
00:22:30.559 --> 00:22:30.799
Yes.
00:22:31.039 --> 00:22:37.119
Over the summer, we had mostly a technical meltdown, which you were describing in the last discussion about uh challenges over the summer.
00:22:37.279 --> 00:22:41.200
When tech gets hit, these little power shell companies get hit twice as hard.
00:22:41.359 --> 00:22:44.240
Not because their business is impacted, but because of technical.
00:22:44.319 --> 00:22:50.640
So we had a meltdown among, frankly, a lot of investors who just saw their positions uh really get hit badly.
00:22:50.799 --> 00:22:52.319
The fundamentals are fantastic.
00:22:52.400 --> 00:22:59.119
And what I see as the escape from this, because you know, how do we get out of this endless trap of just when the markets volatile these stocks go down?
00:22:59.279 --> 00:23:02.880
That path is signing long-term contracts with folks like Amazon.
00:23:02.960 --> 00:23:03.920
And that's what they're gonna do.
00:23:04.000 --> 00:23:09.359
So over the next couple of years, they're gonna turn themselves from sort of a former Bitcoin company into a bond.
00:23:09.519 --> 00:23:12.799
And when they do that, they are gonna trade completely different and to be split.
00:23:14.240 --> 00:23:17.599
When you have a long-term lease with Amazon, yeah, with highly productive.
00:23:17.920 --> 00:23:20.000
10-year, 20-year lease, moving to 20.
00:23:20.079 --> 00:23:21.519
I think all future deals are gonna be 20.
00:23:21.759 --> 00:23:23.519
10's too short, 10's too short.
00:23:23.759 --> 00:23:28.799
With that kind of profile, there'll be REITs and they'll trade completely differently from where they are today.
00:23:28.960 --> 00:23:30.079
That's the upside we see.
00:23:30.480 --> 00:23:41.279
Yeah, now you just mentioned before the video there, Bryce, about the demand for power, you know, uh 33 gigawatts through 2028, 72 to 82 through 2029.
00:23:41.440 --> 00:23:49.839
If you take New York, um the the interview there talks about 7 to 10 gigawatts just required for New York.
00:23:50.000 --> 00:23:52.400
This is the size of at the moment.
00:23:52.559 --> 00:24:03.519
Now, in terms of contracts, when we started covering these companies, um, you know, the durations of the contracts were sort of 10 to 12 years with potentially a couple of five-year extensions.
00:24:03.920 --> 00:24:14.799
As we've seen that mature, we're now seeing effectively a minimum of 15 years as the base period there, with potentially two or three five-year periods on top of that.
00:24:14.960 --> 00:24:24.480
So, anywhere from 15 to uh uh 30 years, and in some instances now we're seeing 20 years as the base period as well.
00:24:24.960 --> 00:24:45.039
When you start looking at uh these contracts, and you know, by the time you've got three or four of these contracts and they're all driving the revenue for a long period of time, these products will effectively become REITs, uh real estate investment trusts, uh, you know, long contracted, um, you know, steady state revenues.
00:24:45.119 --> 00:25:02.240
Yes, they've got um inflation uh adjusters uh year on year, so you know some of these contracts could be going up by you know 3%, 4% annually to cover increases in pricing there, but you've got steady state um margins coming through annually.
00:25:02.400 --> 00:25:20.799
So the market, once that happens, the market will have to give credibility to some of these at the moment, and the fact that we've seen this big pullback since uh June the 22nd in terms of some of these stocks, uh, you can see um you know from you know from the the the like the heat map and etc.
00:25:21.039 --> 00:25:27.039
We you know, where we've shown the pullback just in the last month of 11% on average across all the stocks there.
00:25:27.440 --> 00:25:31.279
Um it's it's it's been a challenging, it's been a challenging three months.
00:25:31.359 --> 00:25:37.599
I know if people look at their portfolios, probably 25-30% uh down since June.
00:25:37.839 --> 00:25:46.319
Um I don't think the market's fully priced in some of these uh some of these contracts, and we've still got a few to be uh to be awarded.
00:25:46.400 --> 00:25:56.720
Um but uh overall um you know it this was showing you even though we're we're we're seeing stocks fall, the positivity around holding some of these HVC AI stocks.
00:25:56.880 --> 00:26:02.079
And in terms of TerraWolf themselves, I mean look at what they've done in the past 18 months there.
00:26:02.160 --> 00:26:25.519
They've got now uh contracts with Core 42 uh fluid stack supported uh backstop by Google and Anthropic, 839 megawatts, further growth of 2.1 gigawatts,$27 billion in contracted revenue, uh, and an average net operating income, an average net operating income of$1.5 billion.
00:26:26.160 --> 00:26:34.319
When you consider uh the Bitcoin mining revenues were probably around about$30 million a quarter, this is a massive step up.
00:26:34.400 --> 00:26:46.559
And in terms of that expansion, in terms of that growth there, they're looking to expand on average 250 to 500 megawatts of capacity signings annually uh from this date.
00:26:46.720 --> 00:26:52.640
And you know, when we've had Patrick on the podcast, um, he's been very transparent about this.
00:26:52.880 --> 00:27:18.000
Uh, and you know, we've had them on for the for more than three years, and they've yet to not reach uh a target or um come good on what they've said um all the way to and go back all the way through when we were when we were we had Paul on and he was buying shares when the share price was in the 50 cent range, and he said, you know, I'm buying these shares because I believe in the company, the company is is not valued at 50 cents.
00:27:18.079 --> 00:27:20.480
I'm buying these shares, and you know, rightly so.
00:27:20.559 --> 00:27:30.160
We saw the share price in June this year reach$30, and we're now seeing the share price as I speak at the moment at$15.
00:27:30.559 --> 00:27:34.000
That's a that's a pullback of 50% just since June there.
00:27:34.240 --> 00:27:37.440
With no real uh reason for that.
00:27:37.599 --> 00:27:50.160
We've seen many uh target prices for terror around the 30 and higher than$30, but today it sits at$15,$27 billion of contracted revenue, one and a half billion manualized NOI.
00:27:50.559 --> 00:27:56.720
I think people need to do a little bit more delving into TerraWolf and look at that share price and think there may be some value here.
00:27:56.880 --> 00:28:04.480
Um, I'm certain looking at it as we speak now to uh to see if there's an opportunity to get some more shares in the next few days from this particular company.
00:28:04.720 --> 00:28:06.640
Yeah, there's definitely a lot of value out there.
00:28:06.720 --> 00:28:13.759
And as we heard in that clip, a lot of the companies we talk about almost guilty by association with this tech sell-off during the summer.
00:28:13.920 --> 00:28:16.400
But you talk about the REIT transition, Anthony.
00:28:16.640 --> 00:28:20.000
Ironic, our company's extremely volatile right now.
00:28:20.160 --> 00:28:28.000
Once they get into that REIT-style model, a lot less volatile as that long-term contracted revenue starts to flow in.
00:28:28.160 --> 00:28:34.640
Now, we've talked a lot about demand so far, we've talked about new models, we've talked about pricing, power shortage.
00:28:34.799 --> 00:28:38.799
As a result of that, we're continuing to see GPU pricing increase.
00:28:38.960 --> 00:28:46.240
Now, we covered Nebius a few weeks back on the channel, AWS doing the exact same thing October 7th.
00:28:46.640 --> 00:28:55.920
They've had their compute prices raise for a full straight quarter, and in Q4, um, they're looking at an average price increase of 15%.
00:28:56.319 --> 00:29:04.000
That's on top of a 20% rise in Q3, a 10% rise in Q2, and a 15% rise for the H200s in Q1.
00:29:04.319 --> 00:29:09.599
So this is cumulative increases here uh significantly over the last four quarters.
00:29:09.759 --> 00:29:18.079
Bear in mind, go back to the first quarter of 2025, most of the machines that are available then, the A100s, H100s, and the H200s, actually went down 20%.
00:29:18.720 --> 00:29:26.799
But this last four quarters, we've seen massive increases, massive demand, massive um changes in the award of contracts.
00:29:26.960 --> 00:29:33.759
So looking at the co-location contracts, which the majority of companies we talk about on a daily basis, they've got co-location contracts.
00:29:33.920 --> 00:29:41.039
The likes of TerraWolf there um have got some big co-location contracts, that 839 megawatts of power.
00:29:41.200 --> 00:29:52.240
Um, we're seeing those rates increase because effectively uh the the the potential clients, the neo clouds and hyperscalers, are actually uh you know getting more revenues from the GPUs.
00:29:52.319 --> 00:30:00.640
So uh the offset of that is is that the the the companies who build the infrastructure are also able to benefit from those increased changes as well.
00:30:00.799 --> 00:30:16.559
However, this is really good for the likes of Iron and for the likes of Corvex and White Fiber and Hive, and also BitDeer who've who've joined and are producing their own uh CSP models.
00:30:16.720 --> 00:30:23.119
Um, this is only going to benefit them you know significantly because they'll get the full benefit of these price increases.
00:30:23.279 --> 00:30:32.400
You've already alluded to the fact that you know that contract with Microsoft which was signed over a year ago, uh the rates at the time there were felt to be you know extremely good.
00:30:32.480 --> 00:30:52.400
It was about 10 million dollars per megawatt, and um you know those contracts now on a three-year basis are you know looking heading towards 25 million dollars per megawatt and on a very short basis of less than 12 months, aiming to between 40 and 50 million dollars a megawatt.
00:30:52.480 --> 00:31:11.279
And with these types of increases, this continues in 2027, it's going to take those numbers to a whole new different level, and we'll start seeing the co-location deals, which went within 12-13 months from 1.4, 1.5 million per megawatt all the way to 2.4, 2.5 million per megawatt.
00:31:11.440 --> 00:31:21.039
That's the the the difference there for co-location, big, big rises, but it potentially here even bigger rises for uh the CSP contracts.
00:31:21.200 --> 00:31:41.920
If you look at uh the new US GPU rates, B300s uh now at$16.14, that's up from$14.04, B200s at$14.20, that's up from$12.35, and the H100s who are now going past the five-year um real uh life period.
00:31:42.079 --> 00:31:51.920
I mean, you know, most people were talking 18 months ago that these GPUs had a five-year life because we hadn't got sufficient historic data to calculate more than that.
00:31:52.079 --> 00:31:57.599
We're now seeing contracts for CSP models signed, not for five years, but for six years.
00:31:57.759 --> 00:32:04.400
And the H100s, uh they're getting close to six dollars per accelerator hour up from five dollars nineteen.
00:32:04.799 --> 00:32:05.759
Six dollars.
00:32:06.000 --> 00:32:18.240
Just to put that into context, I think Iran have got some of the B200s at their at their sites with Microsoft, they're getting two dollars ninety-one for the next five years for those.
00:32:18.400 --> 00:32:28.559
Um, so if these rates are able to stay in this bracket here, you can see that the rates are literally four to five times more than they were 12 months ago.
00:32:28.720 --> 00:32:41.599
And that means future CSP contracts, you know, you won't get the guarantee of you know 60, 14 or 16 dollars, but you'll get significantly higher than$2.91 based on the current rates there.
00:32:41.680 --> 00:32:48.640
So they'll be able to sign up some really good uh deals for the remaining five gigawatts of power that they have available.
00:32:48.960 --> 00:32:56.640
Yeah, and it's interesting to see the differences in hyperscale pricing, which is historically higher versus NeoCloud, and then goes down from there.
00:32:56.720 --> 00:32:59.359
So these are even higher than what we saw from Nebius.
00:32:59.680 --> 00:33:02.559
And again, the question is how high can they go, Anthony?
00:33:02.640 --> 00:33:05.839
The bottleneck continues to be constricted here.
00:33:06.000 --> 00:33:09.759
Now, you mentioned an interesting point: a rising tide lifts all boats.
00:33:09.920 --> 00:33:14.640
We've seen obviously chip prices impact NVIDIA, the chip makers.
00:33:14.720 --> 00:33:17.759
We've seen the CSP model, the colo model impacts.
00:33:17.920 --> 00:33:20.799
We're also seeing it on the DRAM and memory side.
00:33:20.960 --> 00:33:24.799
So we're gonna work into Micron Q4 results here.
00:33:24.960 --> 00:33:27.039
An interesting fiscal year for them.
00:33:27.279 --> 00:33:30.240
These numbers, absolutely monstrous, Anthony.
00:33:30.559 --> 00:33:36.400
Looking to give NVIDIA themselves, the largest company in the world, a run for their money here.
00:33:36.799 --> 00:33:37.759
No, absolutely.
00:33:37.920 --> 00:33:42.720
We are talking um different levels of revenues than we normally talk about.
00:33:42.960 --> 00:33:45.279
So um, let's put this into context.
00:33:45.519 --> 00:34:04.559
You know, uh, Iron, one of the biggest companies that we talk about on a daily basis, there, they've just achieved um and they will by the end of the year a 4 billion annualized revenue based on the fact that um all their 480 megawatts of compute power contracted out, um, and they'll be getting 4 billion of annualized revenue.
00:34:04.640 --> 00:34:06.880
So that's a billion dollars every quarter.
00:34:07.119 --> 00:34:21.599
Micron have just released their latest quarter and they achieved$54 billion of revenue, um, which beat the estimate of just over$51 billion quite significantly, about 6% higher.
00:34:21.840 --> 00:34:24.559
Adjusted EPS,$33.42.
00:34:24.639 --> 00:34:27.440
That was up from$31 is the estimate there.
00:34:27.599 --> 00:34:32.559
The cloud memory revenue represents$16.2 billion.
00:34:32.960 --> 00:34:48.800
That's up from$15 billion as the estimate, and here's where it gets really interesting: the Q1 adjusted revenue outlook now is between 60 and 63 billion, and that's higher than the 56 billion estimate that was previously provided.
00:34:48.960 --> 00:35:05.679
Um, we're talking uh, you know, not many companies can compete with the likes of Nvidia, but here's one that not only can compete, I was listening to an interview earlier today on CNBC, and that the potential is that this company will drive more revenues than Nvidia.
00:35:05.920 --> 00:35:17.920
And uh, just when you think the largest company in the world is NVIDIA, um, Micron only recently IPO'd, and they've gone into that sort of like I'm gonna say we've got like a Mag 10 now.
00:35:18.000 --> 00:35:43.440
We used to have a Mag 7, but I think there's probably about 10, 11 big, big companies, and the majority of these companies are all in this sector at the moment that we talk about, this AI HPC sector, and they're gonna grow exponentially as we start seeing the sheer demand and the revenues that will follow for people wanting to use AI on a on a on a daily basis.
00:35:43.599 --> 00:36:04.159
We are using at the moment in terms of the tip of the tip of the iceberg, but once people really see the benefits and how they can help, how it can help them to drive not only just their personal daily lives but their business lives as well, uh, that's when the revenues will start coming in because tokenization, the usage of these types of systems will bring in revenues.
00:36:04.239 --> 00:36:14.559
Now, in terms of um the um the net income, so we always like to know uh you know it's alright saying the revenues are 54 billion, but what's the net income?
00:36:14.800 --> 00:36:19.599
Net income jumped to 37.7 billion dollars.
00:36:19.760 --> 00:36:31.360
Uh, if you go back 12 months ago, that figure was 3.2 billion price, a staggering swing driven by the memory shortage and the spike uh in pricing over that period.
00:36:31.519 --> 00:36:43.760
Now, that guidance I've talked about there, the 61.5 billion is way ahead of this of the consensus with adjusted EPS guidance of$38.15 uh versus the$34.40 expected.
00:36:44.000 --> 00:37:11.119
And one of the reasons that this company is able to grow revenues that big is it's committing$250 billion to the new HBM campuses, which include the clay New York site that broke in January and Boise, Idaho coming online next year as Micron will work with its uh competitor NVIDIA on the industry's first custom HBM implementation.
00:37:11.760 --> 00:37:14.639
Um, it's just it's just phenomenal numbers.
00:37:14.719 --> 00:37:25.199
Uh, if you think about the fiscal year, because this was their not only their Q4, but their fiscal year for 2026, they surpassed a hundred billion dollars.
00:37:25.440 --> 00:37:29.599
Um, again, this number will be significantly beaten next year.
00:37:29.679 --> 00:37:41.280
This number will be close to 300 billion, I would imagine, by the end of next year, and that puts them clearly um in the sort of like uh in the area of where where NVIDIA are.
00:37:41.360 --> 00:37:52.079
So both these companies benefiting significantly from the demand of AI because they both provide the much needed memory and the chips support in that particular sector.
00:37:52.320 --> 00:37:55.760
Yeah, and we've got the all-time view here up of Micron.
00:37:55.840 --> 00:38:04.400
You can see just in the past year, up 540%, almost similar to Bitcoin in terms of all-time return here, Anthony.
00:38:04.480 --> 00:38:08.960
And the final piece on Micron there, Anthony, just a little interesting fact.
00:38:09.119 --> 00:38:12.000
Their fiscal year actually ends September 3rd.
00:38:12.079 --> 00:38:14.880
So the first day of 27 is September 4th.
00:38:14.960 --> 00:38:22.079
So if you're wondering how they get their full year earnings out this quick, uh, essentially they had 27 days to do it.
00:38:22.239 --> 00:38:26.559
Now, following obviously big earnings like this, we like to talk price targets.
00:38:26.719 --> 00:38:28.960
We've got a number of targets on MU.
00:38:29.119 --> 00:38:36.159
We have one for NVIDIA, one for Core Weave, just to show the overall consensus or excitement in this space.
00:38:36.639 --> 00:38:42.480
Yeah, I'm looking at the Cancer Fitzgerald um uh target for uh for MU at the moment.
00:38:42.639 --> 00:38:58.159
Uh they've they're maintaining an overweight uh rating and uh a maintained$2,000 target, which based on the current price gives it a projected upside of 87 uh point seven percent.
00:38:58.320 --> 00:39:06.159
Um and and you know the key the key takeaways memory and storage suppliers expect to remain under supply into 27-28.
00:39:06.639 --> 00:39:19.119
75% of 2027 output is committed um with 50% committed through to 2030, and the gross margins um are guided at 86.25%.
00:39:20.239 --> 00:39:29.679
Remember, the net income uh margins are probably 60%, so massive gross margins.
00:39:29.760 --> 00:39:31.360
I mean, we talk in the mining space.
00:39:31.440 --> 00:39:39.679
If you're able to achieve you know uh 40 to 50 percent gross margins, then net income margins maybe take most of the rest of that out there.
00:39:39.840 --> 00:39:48.079
Uh, this particular sector is um is one there where there is if you can deliver and execute uh you'll benefit significantly.
00:39:48.239 --> 00:40:05.679
Now, CapEx is expected to be about 55 billion uh during uh 2027 compared to the uh prior 50 billion indication there, and the management themselves see that free cash flow will accelerate from 33 billion um in August in the August quarter.
00:40:05.840 --> 00:40:11.760
So, you know, for for for actual uh Cancer Fitzgerald, this remains one of their one of their top picks.
00:40:12.079 --> 00:40:16.159
Yeah, and we've got a list of all the banks or a majority of them covering MU.
00:40:16.239 --> 00:40:20.880
You can see in every instance, regardless of price target, it's above where we're at now.
00:40:20.960 --> 00:40:23.360
I think the high side around 2200.
00:40:23.519 --> 00:40:27.840
Uh NVIDIA core, we've also got some nice price target updates here, Anthony.
00:40:27.920 --> 00:40:36.159
Again, showing the demand for chips, for memory, for compute, for power, uh, really anyone touching AI.
00:40:36.559 --> 00:40:41.199
Yeah, um, and and and again, uh, you know, this one comes from Rosenblatt.
00:40:41.519 --> 00:40:45.920
Um, they've reiterated a buy rating for Core Weave.
00:40:46.320 --> 00:40:49.599
Uh, remember, we've covered Core Weave significantly during that period.
00:40:49.679 --> 00:40:56.960
They were trying to uh out purchase or acquire Core Scientific, which the shareholders decided to vote against.
00:40:57.039 --> 00:41:10.239
Uh, they thought there was better opportunity in Core Scientific maintaining its um independence from that and and and carrying on and delivering great sites, and uh they've nearly delivered that whole contract to Core Weave now.
00:41:10.400 --> 00:41:24.400
If you look at the uh the price target, that is still a maintained$250 target, which you think of the upside from today, uh, that's 186% upside from the current share price, hovering around$87.
00:41:25.519 --> 00:41:30.239
Now, again, strong demand supporting this price increase.
00:41:30.480 --> 00:41:48.480
Um, the recent price pricing in terms of uh what we talked about there uh in uh in terms of GPU pricing, that's rising continually beyond the 25% noted on the August earnings call, and enterprise checks show significant value from all their AI deployments as well.
00:41:48.639 --> 00:41:56.000
So, really uh you know very positive view on Core Weave and a company that we cover uh quite regularly.
00:41:56.239 --> 00:42:14.000
And finally, uh we've got the largest um company in the world, uh, we've got NVIDIA and Cancer Fitzgerald have provided a uh or reiterated a overweight rating on the company there, they're maintaining their$350 price target.
00:42:14.159 --> 00:42:16.639
Looking at the share price, they$228.
00:42:16.960 --> 00:42:19.599
That's a projected upside of over 53%.
00:42:20.559 --> 00:42:26.880
And again, the you know, looking at the robust demand growth across that increasing broad customer base.
00:42:27.039 --> 00:42:35.119
Um, Nvidia enables the Agentic AI across all verticals, the rising share of overall IT CapEx spend with each new system generation.
00:42:35.280 --> 00:42:40.000
Remember, all these companies building these sites, they're going to need GPUs to put in these sites at some point.
00:42:40.159 --> 00:42:54.000
Nvidia there waiting for those calls to to service those deals, and there is clear upside to guided 70% plus for calendar year 27 uh growth in revenue.
00:42:54.239 --> 00:43:01.519
So um, you know, again, like um uh Micron, they're going to have a phenomenal year in 2027.
00:43:01.840 --> 00:43:09.599
Uh, probably two of the companies that deliver the you know the biggest amount of uh revenues and net income that I've seen for such a long time.
00:43:09.840 --> 00:43:13.039
Yeah, definitely uh getting into that hyperscale category.
00:43:13.119 --> 00:43:15.039
So that rounds us out today, you guys.
00:43:15.199 --> 00:43:20.639
As you can see, everything from a fundamental view here we can see is headed in the right direction.
00:43:20.880 --> 00:43:24.000
A number of big catalysts discussed on the show today.
00:43:24.159 --> 00:43:29.440
Tomorrow's Friday, we've got the newsletter going out in typical fashion, a number of features in there.
00:43:29.679 --> 00:43:31.360
Soluna BitDear Partnership.
00:43:31.440 --> 00:43:33.440
We've also got OpenAI Anthropic.
00:43:33.760 --> 00:43:37.039
Make sure you guys stop by the website, sign up for the newsletter.
00:43:37.199 --> 00:43:51.280
And just to kind of end off where we started here, Anthony, this was a tweet in response to the Wall Street Journal uh image we showed earlier talking about AI as a percentage of GDP, and you can see Soluna feels like they're in a pretty good spot.
00:43:51.679 --> 00:44:00.719
Well, they are, and if you look at the electricity strain um already showing up, the surging bills, the crowded out industries, the multi-year grid queues.
00:44:00.960 --> 00:44:13.440
ERCOT has classified 266 megawatts across Soluna's Texas campuses as baseload, while 438,000 megawatts of requests wait in line.
00:44:13.599 --> 00:44:23.840
Um, this is a company that you know a few years ago was only able to sort of like manage one project at a time and had to have a you know a joint venture partner in there to help with the capital funding.
00:44:24.000 --> 00:44:34.639
Now they're at a point in time where they're about to start commencing delivery of a hundred megawatts um HPC AI co-location site at Cati 2.
00:44:34.800 --> 00:44:44.800
Caty one, an 83 megawatt site for uh Bitcoin mining hosting deals, has already filled in literally months.
00:44:44.960 --> 00:44:51.280
Um, every megawatts been taken up, and most of those megawatts have been taken up by BitDear, they've got 35 of the 83.
00:44:51.679 --> 00:45:07.679
They're now focusing their eyes on having one of their two flagship sites in terms of AI being uh being built, and they've got an LOI for I think for 100 megawatts on that particular site, 266 megawatts across all their campuses.
00:45:07.840 --> 00:45:13.039
This is where the work really starts now, and and the outcome is you know is structural, not coincidental.
00:45:13.199 --> 00:45:19.519
The company are building the behind the meter at renewable generation sites on power that would otherwise go to waste.
00:45:19.679 --> 00:45:26.480
They've got their own wind farm, and they're not having to compete for grid capacity that communities depend on.
00:45:26.639 --> 00:45:32.239
They're converting stranded energy into productive compute without even adding to peak demand.
00:45:32.480 --> 00:45:34.800
And the investment is coming regardless.
00:45:34.960 --> 00:45:41.199
The question is now whether infrastructure underneath is it built to last the future?
00:45:41.440 --> 00:45:43.599
And that's the million-dollar question, you guys.
00:45:43.679 --> 00:45:48.559
We went from getting contracts, executing contracts, now looking forward into the future.
00:45:48.719 --> 00:45:53.679
Let us know if you're making any trades today, picking up any shares at these depressed prices.
00:45:54.000 --> 00:45:58.079
Your thoughts on today's discussion, and make sure you check out the Corvex interview.
00:45:58.159 --> 00:46:00.000
We'll be back here tomorrow.