If you bought $10,000 worth of Google

stock just 3 years ago, you'd have over

$30,000 today. And if you invested that

money in Nvidia, you'd have over $75,000

right now. That's because these

companies are powering the entire AI

revolution. But Google and Nvidia are

already worth trillions of dollars. So

in this video, I'll show you three

smaller AI stocks set to win big for the

same reasons. They're becoming the

critical infrastructure for the next

phase of AI. Your time is valuable, so

let's get right into it. First things

first, I'm not here to waste your time.

This is an update on the top three neo

clouds now that they've all reported

earnings. CoreWeave, which has over a

quarter million GPUs and $67 billion in

contracts. Nebius for sovereign AI and

sensitive data, basically the pounds

here of AI infrastructure. And IREN,

which now has over 4.5 gigawatts of

secured power for AI data centers. But

over the last few months, all three of

these young cloud companies cranked

their AI spending even higher, and they

missed revenue or earnings expectations,

which triggered downgrades, double-digit

drawdowns in their stock prices, and

even lawsuits. That's why this video is

a little late. New information keeps

coming out, and I tried to capture it

all in one place for you. Either way,

these are three of the highest risk,

highest reward stocks of the entire AI

era. And by the end of this video,

you'll know which ones are actually

worth the risk, which ones to avoid, and

which one is the best stock to buy after

their massive sell-off. I want to make

the best use of your time, so let's

start by updating what they all have in

common. Their markets, their business

models, and the new risks that just

showed up in their latest earnings.

CoreWeave, Nebius, and IREN are all neo

clouds, which means they sell access to

compute infrastructure optimized for AI

workloads like training, fine-tuning,

and inference. Neo clouds exist because

traditional cloud computing companies

like AWS, Azure, and Google Cloud were

originally built around workloads that

run on CPUs. Think email and web

servers, processing transactions,

database management, and reporting.

Businesses still need those services,

but the general-purpose servers that

they run on are too slow and expensive

for frontier AI workloads. So, at a high

level, Neoclouds are a new class of

cloud company with dense racks of GPUs

connected by high-speed networks and

filled with advanced cooling to squeeze

the most performance out of every watt,

while their software focuses on

coordinating clusters of GPUs for AI

workloads instead of spinning up CPUs to

host websites and databases. All three

companies earn revenue when customers

rent their AI infrastructure, usually

through multi-year contracts. And

they're spending so much money because

their margins depend on how fast they

can acquire land and power, build data

centers, and deploy GPUs, how much they

can keep those GPUs utilized, as well as

their pricing power versus their

competitors, including Amazon,

Microsoft, and Google. Speaking of

which, I pointed out that Google and

Nvidia made investors rich by powering

the AI revolution. Well, the global

Neocloud market is expected to almost

30X in size over the next eight years

and become a trillion-dollar market by

2034. That would be over a 50% compound

annual growth rate for the next eight

years, which is three times more than

the growth of the S&P 500. That's the

market that Coreweave, Nebulous, and

Iren are all competing in. That's why

they have such similar business models

and risks, and that's why all three

underperformed their earnings calls this

past quarter, at least according to the

headlines. At a high level, they're all

spending more money than investors were

expecting on land, power, GPUs, and data

center infrastructure, and all three

basically told the market that 2026 and

2027 will be years of extremely heavy

spending, not maximizing near-term

profits. And on all three earnings

calls, management said that they're in a

race against the clock. They need to

lock in land, power, and hardware to

turn their multi-billion-dollar

pipelines into real revenue before their

competition can. Remember, in a lot of

ways, this is a zero-sum game. Every

piece of land, every gigawatt of power,

every high-end GPU or memory chip that

one company gets is capacity that others

can't have. And losing that race is

another risk they all share, especially

versus the hyperscalers. CoreWeave,

Nebius, and Airen are all still small

companies, which makes their stocks very

volatile, as I'm sure you've noticed by

now. That's why I'm covering all the

risks up front. But, we should also talk

about their huge upside potential, which

is even bigger after their latest

earnings calls. So, now that we've

covered what these companies have in

common, let's dive into each one

individually to see which stock could be

the best investment. But, if you've been

watching this channel for a while, you

know the best investment you can make is

in you, especially if you're trying to

make a big change in your life. Whether

you want to break into a new career and

work from home, or level up your current

job with AI, Careerist has you covered.

Their AI automation bootcamp is no joke.

It's about 7 hours a week for 4 months,

but it's the fastest path I found to go

from where you are now to making

anywhere from 90 to 140,000

a year working from home with no coding

required. Careerist got their start at Y

Combinator, just like Airbnb, Dropbox,

and DoorDash. And today, they have over

a thousand graduates at companies like

Google, Amazon, Meta Platforms,

Salesforce, and Uber. That's because

they don't just train you and leave you

hanging. You'll build real automation

projects and learn to present them to

real employers. They even help you with

your resume, cover letter, and interview

prep so that you actually get the job.

That's why Course Report named them one

of the best online boot camps, and they

have tons of great reviews from their

alumni. Enrollment for the next boot

camp closes soon, and spots are filling

up fast because big companies need tons

of AI and automation professionals right

now. So, if you're tired of being

underpaid, and you want a

recession-proof career in AI that lets

you work from anywhere, invest in

yourself today by using my link in the

description to get up to 30% off this

awesome AI automation boot camp. All

right, let's cover Coreweave's updates

first since they're the current leader

of the pack. Coreweave operates 43 AI

data centers with over 850 MW of active

power generating revenue right now, and

over 3.1 GW of total contracted power.

Nvidia Blackwell racks use around 120 kW

each, so Coreweave has enough contracted

capacity to power almost 26,000 racks,

or almost 1.9 million GPUs across North

America and Europe. And Coreweave sells

access to those GPUs either as bare

metal chips or full-stack instances

managed through a mission control pane.

Coreweave stack has five layers: the

GPUs themselves, AI-optimized data

storage and networking, software to

automatically manage and scale AI apps

and workloads across many servers,

runtime acceleration tools to speed up

training and inference, and

application-level tools to build, test,

deploy, and monitor AI models and

agents. These full-stack instances make

their platform much more sticky. It

gives them higher margins and more

pricing power versus their competitors.

As of their latest earnings call,

Coreweave's quarterly revenue came in at

$1.6 billion,

which is up 110% year-over-year. Their

full-year 2025 revenue was $5.1 billion,

up a whopping 168%

from the year before. But like I say

every time I cover Coreweave, the

special thing about them isn't their

revenue growth. It's their revenue

backlog, which just hit $66.8 billion

after growing by year. And not only is

their backlog more than four times

bigger than it was a year ago, but their

average contract length is now five

years instead of four, which is an

insane amount of revenue visibility for

a company this young. Most of that

pipeline comes

from a handful of huge AI customers.

OpenAI has up to $22.4 billion worth of

contracts with Coreweave across several

expansions, making it one of the single

largest infrastructure deals in the

entire AI ecosystem. Meta Platforms has

another $14.2 billion agreement to

access Coreweave's Nvidia Blackwall

Ultra systems through 2031. And they

have a $6.3 billion cloud capacity order

from Nvidia themselves, where they'll

purchase any unsold cloud capacity from

Coreweave through April of 2032. By the

way, Nvidia recently invested another $2

billion into Coreweave at $87.20

per share as part of a bigger deal to

build more than 5 GW of AI factories by

2030. At the time of this recording,

Coreweave is trading at $75 per share,

or more than 10% cheaper than what

Nvidia just paid for the stock. That

aside, management is guiding for $12 to

$13 billion in revenue for 2026, which

would be another 140%

year-over-year jump. The reason the

stock dropped after earnings is because

Coreweave plans to spend $30 to $35

billion in CapEx this year alone, and

they already have more than $21 billion

of debt on their balance sheet. So,

investors are worried that their backlog

doesn't justify all of that upfront

spending. On top of that, there are

several shareholder lawsuits claiming

that Coreweave oversold how fast they

could turn their massive backlog into

real revenues. Full disclaimer, I'm not

a lawyer, and this is just my personal,

non-professional opinion on what's going

on here. From what I can tell, these are

all civil securities cases, not criminal

fraud charges. Basically, a bunch of law

firms want to represent shareholders who

bought CoreWeave stock after the IPO and

then lost money when the stock dropped

due to guidance cuts, data center

delays, and the recent sell-off around

earnings. The core argument is that

CoreWeave didn't fully spell out all the

risks associated with their data center

strategy and supply chain for investors.

This kind of lawsuit is very common for

any stock that has a crazy IPO and then

falls back to earth. And the cases are

almost always about the same thing. Big

forward-looking statements about growth

that later look overly optimistic when

reality forces a company's leadership to

lower guidance. Usually, these kinds of

lawsuits get dismissed or settled for

small amounts of money because the legal

bar is so high. The courts would have to

find that specific executives either

knowingly or recklessly misled

investors, not just that they were

bullish and then wrong. If CoreWeave's

disclosures have all the right language

and there's no clear evidence that

management intentionally hid major

problems, the entire case falls apart.

Personally, as a CoreWeave shareholder

myself, who listens to all the earnings

calls, I didn't feel blindsided by

anything that's come out so far. So, I

see these lawsuits as noise rather than

something that changes my long-term

thesis on the stock, which is still very

positive given CoreWeave's massive $67

billion backlog, their full-stack AI

cloud platform, the huge amount of

contracted power they have, and their

massive data center footprint, all of

which make them a clear long-term winner

in this quickly growing market. All

right, let's cover Nebius next. Nebius

is a neo cloud focused on sovereign AI

and regulated industries like healthcare

and financial services. Kind of like

Palantir, but for AI infrastructure. As

of their latest earnings, Nebius made

about $228 million in quarterly revenue.

That's much smaller than CoreWeave, but

it's also growing a lot faster, up 547%

year-over-year. With their core AI cloud

revenue growing by over 800%. Nebius hit

a $1.25 billion annual run rate at the

end of last year, which beat their own

guidance, and management is still

targeting a $7 to $9 billion run rate by

the end of 2026 with $3.2 billion in

full-year revenue. That's another six or

seven X jump in a single year. And just

like CoreWeave, the plan to get there is

a mix of massive hyperscaler contracts

and very aggressive CapEx. Nebius has a

$17.4 billion

five-year AI infrastructure deal with

Microsoft and a separate $3 billion

five-year deal with Meta. Just those two

contracts are roughly $20 billion

of committed AI infrastructure revenue,

which means that Nebius' capacity

is effectively sold out through early

2026. That's exactly why they're

spending around $18 billion

in CapEx to grow that capacity. By the

end of 2025, Nebius had around 170

megawatts of active power, well above

their original 100 megawatt target. They

deployed five new locations and secured

nine additional sites, ending the year

with 16 sites and more than two

gigawatts of contracted power locked in

for their future growth. But Nebius is

doing more than just renting GPUs to

regulated industries. In 2025, they

acquired AVride, an autonomous vehicle

company, and then teamed up with Uber on

up to $375 million of investments and

contracts for robo-taxis and delivery

robots. And just last month, they agreed

to acquire an agentic search startup

called Tavoli for around $275 million.

That gives Nebius search infrastructure

specifically for AI agents in areas like

coding and financial trading, which will

bring hundreds of thousands of

developers into their ecosystem. Nebius'

management is clear about their goal

here. They want to be a full platform

for building and running AI products and

agents, the infrastructure, the software

stack, and now A genetic search. So, big

customers don't have to stitch those

pieces together from multiple vendors.

For investors, Nebius looks a lot like

CoreWeave, but starting smaller and

growing faster with a clear path to

billions in recurring revenue and extra

upside from businesses like AV Ride and

Tavoli. The risks are even more

pronounced here. $18 billion in CapEx

and a schedule that can't afford to slip

on data center construction or on power

delivery, which still needs additional

financing even with Nebius' strong cash

balance and customer prepayments. But,

as an investor, I think that Nebius

still brings a lot to the table. A

strong focus on governments, banks, and

healthcare, plus real optionality in

robo taxis and A genetic search. If they

can execute on everything they're

building, this could be one of the

fastest-growing and most diversified AI

infrastructure stocks to buy for the

long-term. And that brings us to IREN,

the highest-risk, highest-reward stock

of the three. IREN's latest quarter

looked bad on the surface. Revenue came

in at $185 million,

which missed estimates by almost 20% and

is down 23% quarter-over-quarter. Like I

say every time I cover IREN, that

decline is due to them shifting capacity

away from Bitcoin mining and towards AI.

Bitcoin mining revenue fell by about 28%

as IREN lowered their own mining output

and Bitcoin prices fell over the

quarter. IREN posted a net loss of about

$155 million,

but that's mostly due to $219 million of

non-cash items related to converting

their debt to equity and depreciating

their Bitcoin mining infrastructure, all

of which I covered in my most recent

Iran video. If we remove those one-time

non-cash expenses, their adjusted EBITDA

was actually a positive $75 million

at about 41% margins. So, the business

isn't actually in any trouble, but their

AI cloud revenue jumped to $17.3

million,

which is up 137%

quarter over quarter and now accounts

for almost 10% of their total revenue.

The biggest highlight from their

earnings is that Iran now has more than

4.5 gigawatts of secured, grid-connected

power, around 50% more than they had a

quarter ago. And they're still expecting

to hit a run rate of $3.4 billion of AI

cloud revenue by the end of 2026. That

would be a 7x revenue increase year over

year, and that would still only use

about 10% of their total power capacity,

which means they can keep signing many

more multi-billion dollar deals as that

power capacity comes online. But, right

as I was recording this video, Iran

announced that it signed contracts for

over 50,000 additional Nvidia Blackwell

Ultra GPUs, bringing their total target

GPU fleet to 150,000,

with deployments happening through the

second half of 2026. That added capacity

can support over $3.7 billion

in annualized AI cloud revenue, which is

slightly up from the $3.4 billion target

that I just mentioned. Iran says that

they've now secured $9.3 billion in

total funding over the past 8 months to

support their AI pivot. But, now they

expect to finance another $3.5 billion

of capex for those additional 50,000

GPUs and the related infrastructure. The

reason the stock is down is because of

the $6 billion at the market equity

program they also just announced, which

could dilute shareholders by about 140

million new shares or 42% of their

current float. Just to be clear, that

doesn't mean that shareholders are about

to get diluted by 42%. They just

registered for the ability to sell new

shares over time at market prices

without having to do a big offering all

at once. Nothing says that they have to

use the full amount, and the higher the

stock price, the fewer shares that they

need to sell to raise the money they

need. As a long-term investor, I think

this latest announcement proves that

IREN is actually executing on their AI

pivot. They locked in 50,000 more GPUs,

and they found new funding pads as they

keep ramping down their Bitcoin

revenues. They have a fully contracted

$9.7 billion deal with Microsoft over

4.5 gigawatts of secured power, and a

concrete plan to scale to 150,000 GPUs

and $3.7 billion in ARR by the end of

this year. If they can hit their

milestones on time, the last few

quarters will look like noise. But if

they slip on construction, on power

delivery, or on financing, the stock

will stay extremely volatile. That's why

I think they're the highest risk,

highest reward investment of the three.

All right, before we can decide which of

these stocks is the best buy right now,

here's an updated table summarizing

everything I've covered. Just like last

time, I built this table myself by

pulling numbers from each company's

earnings, and I tried to make every row

as apples-to-apples as I could. But

remember, all three companies have

different fiscal calendars, different

contract lengths with hyperscalers, and

they ramped their revenues and backlogs

from very different starting points. The

metrics I use, like number of GPUs and

number of data centers, are all

approximations that depend on what

actual GPUs they end up deploying, the

amount of cooling they use, and how you

count individual facilities versus

larger campuses. So, this table is a

good way to compare their size, but

these aren't official numbers. And now

that we have all that context, we can

finally answer the big question. Which

of these three neo clouds is the best

investment right now? And if you feel

I've earned it, consider hitting the

like button and subscribing to the

channel. That really helps me out and it

lets me know to make more content like

this. Thanks. Now, let's compare these

three stocks. In my opinion, CoreWeave

is still the clear winner of the pack.

They have the biggest revenues, the

deepest backlog, the most GPUs, the

broadest data center footprint, and

Nvidia directly backing multiple

gigawatts of their buildout. If you're a

growth investor who wants a company to

prove their product market fit and have

huge partners derisking their spending,

CoreWeave is still the company for you.

Nebius is quickly growing into the go-to

neo cloud for sovereign and regulated

AI. Smaller revenue today, but now

already over a billion-dollar run rate

with multi-year, multi-billion contracts

with Microsoft and Meta. So, if you're

already all right with the risks and the

volatility, but you want to own the

Palantir of AI infrastructure, Nebius is

the way to go, especially with the extra

potential upsides from autonomous

vehicles and agentic search. IREN is

still the stock for investors who like

optionality and diversification. In a

world where data centers are power

limited, IREN has 4.5 gigawatts of

secured power, giving them a ton of

flexibility in how they monetize their

infrastructure over time. Personally,

I'm still happy dollar cost averaging

into all three because I believe that

this capital-intensive buildout is

laying the foundation for the entire AI

revolution. If I had to pick a winner

now based only on the current price and

scale, I'd still go with IREN since they

have the most power secured while also

being the smallest of the three. But, I

still think that CoreWeave is the best

stock for newer portfolios since they

already have their scale and Nvidia as a

serious partner, giving them the best

risk-to-reward ratio of the three. Let

me know which of these stocks you're

buying below or if you want me to make

another deep dive video on any of them.

And if you want to see even more sites

behind the stocks, check out this video

next. Either way, thanks for watching

and until next time, this is ticker

symbol U. My name is Alex reminding you

that the best investment you can make

is in you.