Marbles Profile
Marbles

@marblesats

Followers
1K
Following
3K
Media
11
Statuses
518

Rich to poor people, poor AF to rich people, shit-poster, $IREN, Bitcoin, stonks

Joined February 2017
Don't wanna be here? Send us removal request.
@marblesats
Marbles
20 days
I read somewhere that if we were to service all of the consumer inference demand and exclude energy required for AI-training, we're still at an enormous gigawatt shortfall which implies this supply / demand crunch persists longer than everyone thinks. This is a pretty interesting
drive.google.com
@DratchCap
Matt Dratch
1 month
Recently spent time updating an older analysis on where AI demand is actually going and came away still thinking we’re massively short compute (~8–50x short) on consumer inference alone. Big range (future is humbling), but even the low end makes the point. I dropped a link to
13
15
148
@marblesats
Marbles
7 days
I don't think people understand how consequential this is. Musk has been brute-forcing his way to construct datacenters faster than then anyone else in the world. Turns out that while you can brute force your to physically build out the datacenter, you cannot brute force your
@edge_of_power
Edge Of Power
7 days
$IREN $CIFR ‼️‼️ The ban on gas turbines for xAI provides a massive tailwind for $IREN and its green generation. Elon Musk’s xAI won’t be able to use natural gas-burning turbines to build out its future data centers like it did in Memphis, following an EPA rule update this week.
18
34
259
@marblesats
Marbles
10 days
Look. When a company won’t stop talking about its pipeline of megawatts, it’s usually because the rest of the story isn’t ready yet. I don’t make the rules.
0
0
21
@bitcoinbutcher1
₿itcoin ₿utcher 🥩 🐑 🐷
11 days
Those who say vs those who do $iren @Agrippa_Inv is back with more alpha via his @SubstackInc 🖇️👇 My favorite part of his deep dive are comments related to construction prowess related to a conversation with @marblesats on Spaces and supported by @FransBakker9812 satellite
5
14
126
@marblesats
Marbles
11 days
Today, January 1st, 2027. CoreWeave: $20B valuation. Nebius: still securing power. Data One: casually reinvented society again. Bitdeer: infinite megawatts, zero customers. Iren: $70B market cap. The market is very rational.
5
3
25
@marblesats
Marbles
12 days
Some companies talk about securing land and power to build data centers. Others talk about tearing down data centers and erecting new ones where power already exists. If you don’t understand why that difference matters, you’re missing the point.
6
1
118
@marblesats
Marbles
12 days
With each Horizon milestone $IREN delivers, the business steadily de-risks: the execution discount compresses, hyperscalers take notice, and the divide widens between those who can truly deliver at scale and those who merely control megawatts. Still early innings.
@FransBakker9812
Frans Bakker
12 days
$IREN Childress 1/9/2026 HD Sometimes, we share our OF content with the public. Please enjoy this satellite image from last Friday. We can clearly see the first concrete slab of Horizon 3 being poured. We can also see the earthworks of Horizon 4 being prepared. We think they
3
9
61
@marblesats
Marbles
16 days
Yes, I am $NBIS and I want to announce the BTM power is so wonderful and innovative that it is far better than grid-connected power in every possible way. Secured, grid-connect power IS NOT VALUABLE. This is why $IREN, a company with primarily grid-connected power will 100%
@RJCcapital
RJC
16 days
$NBIS - EVIDENCE SUGGESTS VINELAND SITE MAY BE FIRST OF ITS KIND FULLY OFF GRID HYPER SCALE DATA CENTER Evidence points to the “systemic innovation” teased for months at the Vineland, New Jersey data center being a large-scale, fully off-grid power solution, built as a
1
0
7
@marblesats
Marbles
2 months
"Building the datacenter is hard, running the datacenter is easy". 😂😂😂😂😂 https://t.co/7xJOI9TPta This interview is arguably worse: https://t.co/U27Qp5u02d Think about this statement for a second. What does this say about $CRWV's role in the AI compute ecosystem? Are
1
0
8
@marblesats
Marbles
3 months
Scare people shitless by saying that it’s inevitable to overbuild capacity for AI. Continues to go on a never before seen run of CapEx investment for Open AI. There are levels to this game.
@slow_developer
Haider.
3 months
Sam Altman says an AI glut is inevitable, maybe in 2-3 years, maybe 5-6, and likely more than once Costs have fallen 40x per year per unit of intelligence, which strains infrastructure "if breakthroughs continue, personal AGIs could run on laptops" We expect demand to grow,
0
0
4
@marblesats
Marbles
4 months
Not only does $EQIX not have access to contiguous hundred of megawatts, they also cannot simply retrofit all of their CPU-configured sites to plug-in GPU's at high enough rack density. Not only that, none of these traditional datacenter companies had enough foresight to
@rftylerpage
Tyler Page
4 months
So, for the past year, “industry experts” told me that hyperscalers would never sign up for a data center far outside the Dallas metroplex…and yet, in our conversations, those tenants told me that they needed hundreds of MW at a single campus….I’m in an Uber to DFW and drove by
0
0
7
@marblesats
Marbles
5 months
$NBIS implied deal goes for 9 MILLION per megawatt per year. The next best deal is $WULF's at 1.85 million per megawatt per year. Either bitcoin miners have been selling their megawatts for cheap or software is that much more valuable. The disconnect here is ridiculous.
@marblesats
Marbles
5 months
$NBIS 17.4 Billion through 2031 with $MSFT implies ~2.7 - 2.9 billion per year. For the NJ datacenter which will 300 megawatts, assuming for the whole site that's around 9 MILLION per megawatt. For context $CORZ deal with $CRWV and $WULF deal with Fluidstack / Google are
15
0
64
@marblesats
Marbles
5 months
$NBIS 17.4 Billion through 2031 with $MSFT implies ~2.7 - 2.9 billion per year. For the NJ datacenter which will 300 megawatts, assuming for the whole site that's around 9 MILLION per megawatt. For context $CORZ deal with $CRWV and $WULF deal with Fluidstack / Google are
11
6
74
@marblesats
Marbles
5 months
Lol. $IREN
3
2
82
@marblesats
Marbles
5 months
$CRWV's CEO is hyping the potential benefits of acquiring Core Scientific, emphasizing the power of vertical integration. In doing so, he’s basically making the case for why $IREN has a structural advantage. Funny.
@mvcinvesting
M. V. Cunha
5 months
$CRWV's CEO is hyping the potential benefits of acquiring Core Scientific, emphasizing the power of vertical integration. In doing so, he’s basically making the case for why $NBIS has a structural advantage. Funny.
1
0
30
@marblesats
Marbles
6 months
Here's how I made my first million, all before age 30. I grinded my fiat job. I read The Bitcoin Standard. I read Broken Money. I bought bitcoin. I doubled down on bitcoin. I listen to pretty much every @mikealfred space. I read every @FransBakker9812 , @TheKamaHsutra ,
@SherryYanJiang
Sherry Jiang
6 months
here’s how i made my first million by 30: - joined big tech early + stacked stock options (rip good ol’ days) - bought way too much crypto for my own good - bought high-growth stocks + actually HELD (tesla, palantir) - never held cash for long (inflation is theft lol) - said yes
7
0
24
@marblesats
Marbles
6 months
Bitcoin miners are shit businesses. $IREN is a bitcoin miner. Therefore, $IREN is a shit business. With 86 million dollars of monthly revenue, $IREN is a shit business. With industry-leading 75%+ gross margins, $IREN is a shit business. With an industry-leading net profit
6
3
89
@marblesats
Marbles
6 months
The $CORZ crowd is right. If $IREN hasn't gotten a deal by now, then $IREN will never get a deal in the foreseeable future. Power is an overrated asset, and time-to-value of power is completely overvalued by these $IREN goons. I have been mislead by $IREN analysts, and should
41
7
157
@marblesats
Marbles
6 months
Today I sold all of my shares of $IREN. Every. Single. Share. Here's why: 1. We are seeing bearish RSI divergence, suggesting momentum is weakening despite rising price — which is a huge warning sign for a possible reversal or consolidation. 2. The chart looks extremely weak.
45
10
177