Ray.NebiusMaxxing 🇸🇬 retweeted
The market is treating a Dem sweep as an existential threat to AI infrastructure. It’s not.
Data centres can (and already are) being built outside the US. Hyperscalers have major footprints and active expansions across Europe, Asia, the Middle East, Canada and LatAm. Power, speed of permitting and cost drive site selection more than pure US policy.
If domestic buildouts slow, capital will simply reallocate. Global AI compute demand doesn’t disappear because Northern Virginia or Texas gets harder to permit. Latency-sensitive training prefers concentrated clusters, but inference, regional capacity and a large share of workloads work fine abroad.
The nutty part is the indiscriminate selling of the whole growth stock complex. High-quality, FCF-positive AI names and infrastructure software companies with multi-year visibility are getting lumped in with pure high-beta, negative-FCF builders that actually do have financing and execution risk.
Policy friction in one country ≠ the end of the AI buildout.
International optionality is real.
RT @TopTechStocksUS: $NBIS There is a lot of talk around about the need of appointing an AI representative to squash the negative rumours t…
$nbis excellent overview of the company.
$NBIS is diversified in at least 7 ways.
If you can count all 7 you will sleep well in times like these.
Let's see:
1. Across the stack
GPU cloud → managed Kubernetes & Slurm → Token Factory inference → post-training (Eigen AI) → GPU snapshots (Inferize) → search for agents (Tavily) → sandboxes.
2. Across countries
Finland, France, Iceland, UK, Netherlands, Israel, US. Saudi Arabia next (NEOM, from Q4 2027). Partners now selling in Türkiye, the UAE and Saudi, plus first customers in Asia-Pacific.
3. Across customers
Hyperscalers (Microsoft up to $19.4B, Meta up to $27B), AI labs, enterprises, startups and universities. Plus distributors like TD Synnex.
4. Across pricing
Multi-year contracts, on-demand, dynamic spot, preemptible, per-token.
5. Across hardware
Hopper → Blackwell → Vera Rubin, plus the first cloud on Nvidia's Groq 3 LPX for ultra-fast inference.
6. Across how it builds
Owned data centers, leases, self-builds, and an asset-light model where partners bring the capital.
7. Beyond the AI cloud
→ Avride: robotaxis with Uber and Hyundai in Dallas, delivery robots for Uber Eats. Uber and Nebius committed up to $375M to it last year.
→ ClickHouse: a reported ~28% stake in a database company valued at $15B in January. Roughly $4B on paper.
→ Toloka (AI training data, backed by Bezos Expeditions) and TripleTen (tech education).
So less dependence on any single customer, country or chip generation.
And the stakes double as a funding source: Nebius has framed ClickHouse as an asset to help fund the AI buildout.
Ray.NebiusMaxxing 🇸🇬 retweeted
$NBIS $NVDA $CRWV $MSFT
Modal CEO @bernhardsson gave the clearest explanation of AI margins I've heard.
"Almost no company today has software margins anymore."
The reason is "almost topological":
"You have applications, you have models, you have infrastructure, you have cloud, you have Nvidia. When a customer pays a dollar at the top of the stack, it has to flow through a lot of different layers, and at every layer there's going to be maybe 20%, 30% gross margins."
His own long-term target: 40-60%. "We're never going to get to software margins."
ELI16: every middleman between your $1 and the GPU takes a cut.
Now look at $NBIS. Nebius designs its own servers, writes its own firmware, runs its own cloud, sells its own tokens through Token Factory, and owns some of its data centers.
The more layers you own, the fewer cuts you pay.
Source: youtube.com/watch?v=1dwyL6Iz…
Ray.NebiusMaxxing 🇸🇬 retweeted
$NBIS EU-West2 paying a whopping $13.10/hr in spot price today. This is the highest I have seen so far 🤯🤯
Ray.NebiusMaxxing 🇸🇬 retweeted
For the record, I don’t consume any of Citrini, SemiAnalysis, and the likes. I’m very protective of my clarity of thought.
Ray.NebiusMaxxing 🇸🇬 retweeted
My guess is all of the Democratic anti Data Center and AI rhetoric will amount to very little. AI is too important to the economy and slowing down won't slow AI it will only let China win.
Every Presidential candidate over the last 20 years has campaigned on ending capital gains treatment for carry and yet it never happens. This is because most of the financial services world is Democratic and at the end of the day they kill it.
Same thing is true in AI. Outside of SPCX almost every company in the AI world is full of Democratic donors.
It is also true that while data centers are not popular they are very low on the list of voter priorities....
Ray.NebiusMaxxing 🇸🇬 retweeted
President of $META:
"We did have a governor that won last night who did make data centers a big part of their platform say, 'As soon as we get past the inauguration, let's meet because we don't want to lose this investment in our town. We don't want these jobs to go elsewhere,'"
Ray.NebiusMaxxing 🇸🇬 retweeted
Citi Research on $CRWV and $NBIS
$CRWV TP: $128
$NBIS TP: $265
They continue to expect neoclouds (at least the bigger ones) to print due to the favourable demand vs supply imbalance. Compute demand will continue being strong, the Dems winning the mid-terms doesn't mean demand for AI suddenly goes to zero. Short-term noise, long term clarity.
Ray.NebiusMaxxing 🇸🇬 retweeted
Let me tell you why the Democrats won.
Gas is $6.50 in my town. That hits every household. Contractors won't go to Home Depot because they're afraid of ICE. Those are the real issues people voted on. Not data centers.
Both parties talked about slowing data center growth while reviewing the backlog. They didn't say stop. They said don't drain the water. Don't raise electricity prices on consumers. That's why data centers are already moving off the grid and away from water cooling.
Wall Street wants you to believe this election was about data center growth going to zero. It wasn't. It was about gas prices and kitchen table economics and immigration.
If you're selling your tech because of that narrative, you're handing your shares to Wall Street at a discount. Six months from now the same sell-side analysts will tell you how valuable Nvidia, compute, and memory are again. And you'll buy back at higher prices.
Don't let Wall Street use an election to shake you out of a structural trend.
Ray.NebiusMaxxing 🇸🇬 retweeted
$NBIS: vCluster is helping Nebius expand Token Factory beyond its own data centers.
Nebius made a deliberate architectural choice: run Token Factory as an elastic platform that can place inference workloads on its own fleet or on qualified third-party GPU capacity, and treat both as the same platform.
Owned data centers stay the foundation. External capacity augments them, absorbing demand as it appears and following it into the regions where it appears.
The result is a Token Factory that grows with the market rather than with the construction schedule.
Nebius proved the model with its own experiments first, then began evaluating bare metal providers to scale it, and brought vCluster Labs in for that step.
The reason matters: third-party capacity comes with someone else's platform attached. Each provider arrives with its own Kubernetes distribution and version, its own observability, backups, security posture, and deployment tooling.
Standardizing that layer before the provider count grew is what kept Token Factory a single platform rather than a set of integrations behind a shared API.
Currently, Token Factory lists 4 external subprocessors across 3 regions:
- RunPod (Iceland)
- BoostRun and Shadeform (US)
- Axe Compute (Canada)
Source: vCluster
🚨 $NBIS adds three new compute clusters
"TF" presuambly stands for Token Factory, with one cluster tagged CA1, likely California.
We know $NBIS doesn't have a facility in California.
So high probability these are asset-light deployments that are now online for managed inference.
In particular, could be for $PLTR enterprise customers.
Ray.NebiusMaxxing 🇸🇬 retweeted
Although $NBIS buildout is completely uncoupled by recent developments in Texas, FUD is overblown even for players with existing interests such as $IREN, $CRWV, $CORZ, $HUT, $WULF etc.
Politicians love to hand wave. They make a lot of noise, and trumpet whatever is in the Overton window because it will get them the most votes from their constituents.
They're called career politicians for a reason.
But let's be clear. Lawmakers will certainly raise the bar for new data center builds. However, this high bar is already being met by quality players in the AI infrastructure space, whether from noise restrictions, community benefits, or emissions profile standpoints.
What these new restrictions will actually do is raise the barrier to entry for hundreds of new "neocloud" players with power, land, and capital but no experience building AI/HPC data centers.
Gina Hinojosa discusses data centers
She said the proposed pause on data centers would apply to new permits while lawmakers work on protections, rather than stopping projects already under construction.
aol.com/articles/big-country…
Don't panic. Short term irrational. Bad sentiments. Remember the SaaS had this attack earlier this year.
This too shall pass.
$nbis
Ray.NebiusMaxxing 🇸🇬 retweeted
1/ Everyone take a deep breath. Texas data centers aren't going anywhere. The Texas governor has very limited authority to stop them. Calls for a pause or moratorium must be approved by TCEQ—the permitting authority of Abbott appointees. Here's how Texas permitting really works👇
