NBIS to $500 Billion
Why I Think Nebius Group Could Be the AWS of AI and Why $500 Billion Is Not Crazy
NBIS is currently 42% of my portfolio. I am long since $25$.
That is not a casual position.
That is a reflection of conviction.
I think the market still misunderstands what Nebius is becoming. Many still look at it as just another AI infrastructure stock, another GPU rental story, another short-term beneficiary of AI hype. I think that view is far too small.
My view is that Nebius is building something much bigger.
It is building the kind of platform layer that could become essential to the AI economy.
That is why I believe NBIS 0.00%↑ has a real path to becoming the AWS of AI and why a $500 billion market cap, while still ambitious, is no longer a ridiculous idea if execution continues.
The market is still thinking too small
Most investors are still stuck in the first phase of the AI trade.
They understand chips.
They understand Nvidia.
They understand that training models requires a lot of compute.
But many still do not fully understand what happens after that.
AI is now moving beyond one-off chatbot usage and into always-on systems, enterprise workloads, agents, reasoning models, inference at scale, and software that needs persistent access to powerful infrastructure. That changes everything.
The next stage of AI is not just about building smarter models.
It is about running them continuously, reliably, globally, and economically.
That means the value shifts toward the cloud and infrastructure layer.
Not a generic cloud.
Specialised AI native cloud.
That is where Nebius comes in.
Why Nebius stands out
Nebius is not trying to be everything for everyone.
That is one of the reasons I like it.
The company is focused on AI native infrastructure. It is building for the exact workloads that are growing the fastest in the market. This matters because AI workloads are very different from traditional cloud workloads.
They need:
High performance GPU clusters
Fast networking and low latency architecture
Efficient power usage
Software orchestration optimized for AI
Scalable data center capacity
Enterprise reliability for massive customers
This is not the same as offering basic cloud storage or generic compute.
Nebius is trying to become the place where serious AI builders go when they need real infrastructure.
That is why the AWS comparison matters.
What I mean by “the AWS of AI”
I do not mean Nebius will literally become Amazon.
I mean it could become one of the foundational infrastructure platforms that developers, startups, enterprises, and large model builders depend on.
That is what made AWS so valuable.
AWS was not just selling servers.
AWS became the operating layer for a huge part of the internet economy.
If AI becomes the defining computing platform of this era, then the companies that become the operating layer for AI could become some of the most valuable businesses in the world.
That is the core thesis.
Nebius does not need to dominate all of cloud.
It just needs to become one of the most important platforms in the AI cloud layer.
If it does that, the upside can be enormous.
The demand side is real
This is one of the most important parts of the story.
A lot of AI names are still selling a dream.
Nebius is already showing signs of real demand.
The company has reported extremely strong annualized run rate revenue growth in its core business. It has also guided toward a steep ramp over the next several quarters and years.
That matters because infrastructure businesses can look expensive before scale and suddenly look cheap once utilization and revenue catch up.
A lot of investors wait for the story to feel safe.
By then, the rerating is often already over.
Here is what stands out to me:
Annualized run rate revenue has already scaled sharply from a much smaller base
Management has guided toward around $1 billion ARR by the end of 2025
Management has also pointed toward a range of roughly $7 billion to $9 billion ARR by the end of 2026
Those are not small aspirations
Those are platform scale ambitions
If Nebius gets anywhere near that path, the market will not value it like a niche supplier.
It will start valuing it like a major AI infrastructure platform.
The mega deals change the conversation
This is where the story becomes much more serious.
Nebius is not just selling to small customers experimenting with AI.
It has announced and discussed major commercial relationships that validate both demand and credibility.
That is critical.
In infrastructure, customer quality matters almost as much as revenue quantity.
When large enterprises and hyperscale level customers trust a provider with meaningful AI workloads, that sends a very strong signal to the market.
The big points here are simple:
Nebius has secured large scale customer agreements
These deals increase visibility into future demand
They help prove that Nebius is relevant at the highest levels of AI infrastructure
They also support the idea that this company is not building in a vacuum
In my opinion, once companies start landing major contracts and proving they can scale into them, the debate changes.
The question is no longer “is this real?”
The question becomes “how big can this get?”
Infrastructure scale is the moat
One reason I am so bullish is because this is not a software story without barriers.
This is a capital intensive buildout.
That can scare some investors, but it is also exactly what can create a moat.
If it were easy, everyone would do it.
Winning in AI infrastructure requires:
Access to high end hardware
Power availability
Data center development
Networking expertise
Deployment speed
Operational execution
Customer trust
Capital discipline
Nebius is actively expanding its data center footprint, including major US capacity growth and additional European expansion.
This matters for two reasons.
First, demand is useless if you cannot serve it.
Second, companies that can bring capacity online fast enough in a constrained market often gain strategic importance very quickly.
In AI infrastructure, speed is not a nice bonus.
It is a competitive weapon.
Why this can become a $500 billion story
Let’s talk about the number everyone focuses on.
$500 billion.
That sounds extreme at first.
But that is only because most people are still anchoring on what Nebius is today, not what it could become if execution works.
Big platform businesses do not grow in a straight line in the public imagination.
They go through phases.
Phase one is disbelief.
Phase two is acceptance.
Phase three is rerating.
Right now, I think Nebius is still in the zone where a lot of the market is underestimating the size of the opportunity.
A path to $500 billion would likely require several things to happen:
Nebius continues scaling revenue aggressively
The company executes well on capacity expansion
Major customers deepen their relationships
The broader market starts viewing Nebius as a platform, not just an equipment and compute story
AI demand keeps compounding across training, inference, and agents
Margins improve as scale and utilization improve
Investors start valuing the company on long duration platform economics
If those pieces come together, then the valuation framework changes dramatically.
At that point, investors are no longer asking whether the company deserves a few extra points of revenue multiple.
They start asking whether this is one of the foundational winners of the AI age.
That is a completely different conversation.
Why the AWS analogy is so powerful
The AWS comparison works because it helps frame how large infrastructure winners can become.
The market often underestimates infrastructure businesses early because they look heavy, expensive, and operationally messy.
But once they become essential, they can become incredibly valuable.
AWS became massive because it turned infrastructure into a platform.
That is exactly why I think Nebius is interesting.
If Nebius becomes the infrastructure platform serious AI companies prefer, then it gains:
Sticky customer relationships
Recurring revenue
Strategic relevance
Better economics over time
More pricing power than the market expects
A stronger competitive position as the ecosystem grows
That is how you go from being seen as a tactical AI trade to being seen as a foundational compounder.
The biggest reason I own it so heavily
There are a lot of AI stocks that can work.
Very few can become central.
That is the difference.
I do not want to only own companies that benefit from AI demand in a broad sense.
I want to own companies that could become part of the architecture of the AI era.
Nebius has that kind of potential.
That is why it is 42% of my portfolio.
Not because it is safe.
Not because there are no risks.
But because if I am right, the upside is not small.
The risks are real
This is a high conviction idea, not a risk free one.
That matters.
There are several things that need to go right:
Execution on massive capex
Fast and efficient deployment of capacity
Continued access to leading hardware
Strong utilization of new infrastructure
Customer concentration risk
Competition from larger and better capitalized players
Market volatility and multiple compression
The challenge of scaling this fast without operational mistakes
This is not a story for investors who want zero uncertainty.
It is a story for investors who understand that some of the biggest winners look messy before they look obvious.
My bottom line
I think the market is still dramatically underestimating what Nebius could become.
This is not just another AI stock.
This is a company trying to build one of the key infrastructure layers of the AI economy.
If it succeeds, the upside is not about a nice move higher.
It is about a full category rerating.
That is why I think
could become the AWS of AI.
And that is why I think a $500 billion market cap is not just hype.
It is a serious long-term possibility.
is 42% of my portfolio for a reason.
I think we are still early.
What do you think is a reasonable timeframe to the 500 Billion? End of decade?