Nvidia just reported their biggest

earnings ever, but below all the

headlines, they're quietly fighting a

war on two fronts, data centers and edge

computing. Both markets just got much

more competitive and they're filled with

great investments, if you know where to

look. My name is Alex and I spent eight

years as an electrical engineer and AI

researcher at MIT, which helped me find

stocks like Nvidia, Micron, Poet and

Iren way before the rest of the market.

And in this video, I'll catch you up on

five major stories that are already

moving markets and changing which stocks

are about to win big. So, let me show

you what just happened and how I'm

investing in it. Your time is valuable,

so let's get right into it. Nvidia

reported earnings a little over a week

ago and it was their best quarter ever

by far, $81.6 billion in revenue, which

is up 85% year-over-year. But while

everyone is focused on the headline

numbers, Nvidia made a small change that

could have big consequences. For the

first time in their 33-year history,

Nvidia will no longer report gaming GPU

sales as a separate segment. Instead,

the graphics cards that built the Nvidia

empire will now be lumped in with PCs

and workstations, game consoles,

robotics and automotive under a single

edge computing segment that represents

less than 8% of their total revenues. By

itself, this doesn't seem like a big

issue. It's just a change in how Nvidia

reports their numbers to reflect the two

big AI markets that they currently

operate in. Data centers and edge

computing. But the problem is that edge

computing is a massive and highly

competitive landscape filled with

companies excited to compete with Nvidia

on their home turf, whether that means

phones, PCs, cars or robots. Companies

like Qualcomm, Apple, ARM, AMD and even

Intel have been dominating different

corners of the edge computing market for

decades and Nvidia's reporting change

just gave them all permission to compare

their strongest business units to

Nvidia's weakest, claim a direct market

share advantage over them, and inflate

their stock prices as a result. Nvidia's

data center business is also facing much

more competition on multiple fronts. On

the CPU side, their CFO Colette Kress

said that the latest Vera CPUs are

expected to bring in around $20 billion

in revenue over the next year. But Arm

and Qualcomm both just shipped their

first data center CPUs as well. And when

it comes to inference, custom AI chips

designed by companies like Google,

Amazon, and Cerebras are much cheaper

and more efficient for specific kinds of

workloads. Let me be clear about what

I'm saying here. My point is not that

Nvidia will outright lose any of these

markets. My point is that their

competition can now directly compare

themselves to Nvidia in Nvidia's weaker

markets whenever it suits them. And

since the mainstream media focuses on

headlines and doesn't really look below

the surface, a lot of stock prices are

about to change. Changes that we can

take advantage of as investors. That's

the focus for the rest of this video.

Let's start with Qualcomm, ticker symbol

QCOM, which is up by around 70% since

the last time I covered them. Just a few

days ago, Qualcomm struck a deal to

supply data center chips to ByteDance,

the company behind TikTok. ByteDance

designed their own custom AI chips, and

Qualcomm is making those designs

production-ready and coordinating with

the Taiwan Semiconductor Manufacturing

Company to build millions of them

through 2026 and 2027. Here's why

investors should care about this deal.

ByteDance is a Chinese company, and

Nvidia isn't allowed to sell their data

center chips to China. But Qualcomm's

deal works because the chips are custom

ASICs, application-specific integrated

circuits built for one specific job. So,

they're not on the restricted export

controls list. So, Qualcomm just found a

way into the largest AI market on Earth.

A market that Nvidia is locked out of.

And this isn't a one-off order. Qualcomm

will be embedded in ByteDance's chip

design and production process, making

them much harder to replace over time

and over a standard supply contract.

Qualcomm's data center plan has two more

parts beyond ByteDance. First, their

Orion CPU directly competes with

Nvidia's Vera chip. And second, their AI

200 and AI 250 inference accelerators

enter the market this year and next year

respectively. That's why Qualcomm stock

had doubled over the last two months.

Look, the stock market is always

changing and it can feel impossible to

sift through all the noise and find the

best stocks. That's why most investors

only find them after they make big

moves. But I've been using GenSpark, the

all-in-one AI workspace sponsoring this

video, to have an AI assistant watch my

stocks for me. I like GenSpark because

it's powerful and easy to use. I just

opened a new project and told the super

agent to build a simple workflow. Pull

my watchlist every day after the close,

scan for any unusual price or volume

moves, and write a short summary of what

changed and why. Then I used AI sheets

to define the rules. Things like flag

anything that moves more than 5% or

trades double the normal volume.

Finally, I made it an automated

workflow. Aftermarket close, GenSpark

runs the checks and sends me a Slack

message with the tickers to pay

attention to. It feels less like a tool

and more like having a real assistant.

That's why a ton of people are already

quietly using it to be more productive.

And that's why GenSpark went from

concept to a $250 million annual run

rate in just 12 months. And right now

they're offering unlimited use of AI

chat and AI image for all paid users in

2026. Unlimited subject to abuse

guardrails. You can try it with free

credits using my link and set up the

same workflow for the stocks you care

about. I'll leave my prompts in the

description below. All right, so the

thesis on Qualcomm stock just changed.

For years, they were priced as a

smartphone chip company with revenues

that rise and fall based on how many

smartphones people buy each year. But

now, they're fighting on the two biggest

fronts of the AI chip war, data centers

and edge computing. Qualcomm's

automotive segment brought in 1.3

billion dollars in revenues last quarter

and grew 38% year over year. Nvidia's

automotive business grew just 6% year

over year and had less than half the

revenue. Hopefully, you're starting to

see what I mean about those headline

comparisons. And investors shouldn't

sleep on edge computing. The global edge

AI market is expected to almost 6x in

size over the next 8 years, which would

be a compound annual growth rate of 24%

through 2034. That's close to twice as

fast as the growth of the S&P 500. But

while edge AI is becoming a bigger

battleground, the data center market

still represents about 2/3 of all AI

accelerator revenues and it's growing

even faster at 28% per year. That means

it's expected to more than 7x in size by

2034. So, every company competing in it

should see serious growth even if their

market share stays the same. Companies

like Arm and Cerebrus. For the last 35

years, Arm let everyone else fight the

chip war while they sat back and

collected royalties from all sides,

Nvidia and AMD, Apple and Qualcomm.

Then, at the end of last quarter, they

announced that AGI CPU, the first

production chip Arm has ever designed,

manufactured, and branded for

themselves. This chip is not a warning

shot. It's a tactical nuke. In a recent

video, I compared Arm's AGI CPU to

Nvidia's Vera CPU and long story short,

Arm's new CPU is more powerful to the

point where data centers need around 40%

less of them to support the same amount

of GPUs. And that's just versus Nvidia.

It has around double the performance per

watt compared to Intel and AMD's chips.

Arm expects to sell over a billion

dollars worth of AGI CPUs in the first

year alone, and hit 15 billion dollars

in annual trip revenue within five

years. The whole company makes less than

five billion dollars a year today. So,

this chip would effectively quadruple

their total revenues by 2031. And one of

the first companies pairing this chip

with their own AI hardware is Cerebrus,

ticker symbol CBRS, which just went

public with the largest US semiconductor

IPO ever. Every chip on Earth gets

stamped out in a large silicon disk

called a wafer, and that wafer gets cut

into hundreds of individual chips.

Cerebrus skips that step entirely and

turns the whole wafer into one massive

chip called the wafer scale engine, or

WSE. Their current generation is the WSE

3. And in my most recent video, I

compared it to Nvidia's Blackwell chips,

since commercial shipments of Vera Rubin

don't start until quarter three. In a

nutshell, Cerebrus's chips are 62 times

bigger. They have 19 times more

transistors, 44 times more AI cores, and

a quarter of the memory, but 2600 times

the memory bandwidth. As a result, this

wafer scale engine can run Meta's Llama

4 Maverick model roughly 2.4 times

faster than the B200. That's because

Nvidia has to move data between multiple

separate chips, across cables, and

through network switches, all of which

adds time to every transfer. While

Cerebrus just moves data across one

giant chip. I'll leave a link to my

videos covering Arm's AGI CPU and

Cerebrus's wafer scale engine below. But

at a high level, their exact speed

advantages depend on the actual

workload. And there are plenty of cases

where Nvidia still wins by large

margins. Not to mention that Nvidia's

CUDA platform has two decades of

software, developer tools, and

infrastructure that every AI team

already relies on. But, the common

thread here is clear. Qualcomm, Arm, and

Cerebras are now directly competing in

Nvidia's market, which wasn't true just

1 year ago. And while every company I've

covered so far uses completely different

architectures, they all have one thing

in common, and that's memory. Just a few

days ago, the only US company making it

crossed a trillion dollars in value.

Here's a few interesting facts about

Micron, ticker symbol MU. I cover the

stock very often, so I'll keep it short

and sweet. Micron is the only US company

that makes high-bandwidth memory for AI

data centers. Its biggest competitors

are SK Hynix and Samsung, both of which

are great companies, but they're based

in South Korea. That means they're more

affected by things like tariffs, trade

wars, and conflicts like the Iran war,

which closed critical supply lines

between the Middle East and Asia. But,

because Micron's in the US, they're not

affected the same way. Micron's

high-bandwidth memory can be found in

Nvidia's Hopper, Blackwell, and Rubin

chips, in AMD's Instinct MI300 and 400

series accelerators, and even in

Google's TPUs. Although SK Hynix has the

larger share of Nvidia's Blackwell

memory. Either way, Micron was already

sold out of high-bandwidth memory for

all of 2026 as of their earnings call a

few months ago. It's kind of hard to

overstate how fast Micron is actually

growing. They reported record revenues

of almost 24 billion dollars last

quarter, which was already up nearly

200% year-over-year. Their gross margins

came in at 75%, which is better than

most software companies. Their net

income grew by almost 20 x and their

earnings per share grew by almost 30 x.

But what's even crazier is their

guidance for next quarter, 40% revenue

growth, another 6% increase in gross

margins, and 57% earnings growth. Not

year over year, but quarter over

quarter. Said another way, Micron will

make more money next quarter than they

made in any full year in the company's

history before 2025. Like I've been

saying for years now, memory is no

longer a commodity. It's a core

component of the AI revolution. Micron

just became my third biggest winner of

all time, only behind Nvidia and Poet

Technologies, and just above Palantir.

This stock skyrocketed by more than 10 x

over the last year and tripled in price

over the last 5 months. And believe it

or not, Micron is still cheap. It trades

at a forward price to earnings ratio of

just 12, while other chip companies

trade anywhere from a 20 to 80 forward

PE, even though Micron's earnings are

growing faster. Maybe the next time I

call a company the next Nvidia, someone

will finally believe me. Speaking of

skyrockets, the space market is about to

have its own chat GPT moment when SpaceX

IPOs. And if you feel I've earned it,

consider hitting the like button and

subscribing to the channel. It really

helps and it lets me know to make more

market recaps like this. Thanks. Now,

let's talk about space stocks. SpaceX is

widely expected to IPO on June 12th with

the ticker symbol SPCX. It's expected to

be valued at close to $2 trillion,

making it the largest IPO in stock

market history by a huge margin. And I

expect every single space stock from

Rocket Lab to ASTS to move up and down

with it, just like AI stocks move with

Nvidia. That's why I want to talk about

Rocket Lab right now, ticker symbol

RKLB. Until SpaceX IPOs, Rocket Lab is

the only vertically integrated

end-to-end space company that's publicly

traded. And the question I get very

often is if this is a stock worth

buying. While most space companies

either build satellites or launch

vehicles and run the missions, Rocket

Lab does all three. Rocket Lab's

Electron rocket is the most frequently

launched small rocket in the world,

carrying small payloads like satellites

and research equipment into low Earth

orbit. It's a dedicated launch vehicle,

which means customers reserve the whole

rocket instead of sharing a ride on a

bigger one. One big thing that investors

might find interesting about the

Electron rocket is its Rutherford

engine, which is the first rocket engine

to have its main components 3D printed.

The combustion chamber, the injectors,

the pumps, and the propellant valves are

all 3D printed via a process called

electron beam melting or EBM, which uses

a high-powered electron beam to fuse

metal powder layer by layer. This cuts

manufacturing time from months to days,

and it dramatically reduces the cost per

engine. It also uses an electric pump to

push propellant into the main combustion

chamber instead of a separate gas

generator, which is a fundamentally

simpler and lighter design. Earlier this

month, Rocket Lab reported their best

quarter ever, $200 million in revenue,

which is up over 60% year over year, and

their backlog just hit $2.2 billion

dollars, up more than 20% quarter over

quarter and 100% year over year. Their

backlog has three major buckets. First

is small satellite launches on their

Electron rocket and medium lift launches

using their Neutron rocket. Second,

defense contracts for hypersonic test

flights and suborbital missions using

modified Electron rockets. And third,

contracts for manufacturing satellites

and spacecraft. About 2/3 of their

revenue actually comes from the space

systems manufacturing contract versus

the 1/3 that comes from launch services.

But space is by far the toughest market

to operate in. Just a few days ago, Blue

Origin's New Glenn rocket exploded

during a ground test here in Florida,

badly damaging the launch pad. No one

was hurt and no satellites were lost,

but their next launch was scrubbed.

Competition lives and dies by these

launches and the entire space sector

sold off the next day. So, can Rocket

Lab actually compete with SpaceX in such

a tough market? Here's a table I made to

help us compare them side by side.

SpaceX is roughly 30 times bigger than

Rocket Lab by revenue and launches eight

times more rockets per year. The two

companies aren't competing for the same

customers right now. SpaceX launches big

payloads on a $74 million rocket. Rocket

Lab launches small satellites on an $8

million rocket with a 3D printed engine.

Different markets, different price

points. But here's where things get

interesting. SpaceX is going public at

roughly 60 times forward revenue while

Rocket Lab trades at 20 times. So, if

you believe SpaceX's valuation is

justified, then Rocket Lab looks cheap

by comparison, even at an $84 billion

market cap. And here's one cool detail

to bring everything full circle. SpaceX

is one of the first customers for

Nvidia's new Vera CPU. So, the same chip

war that we started with, Nvidia

fighting on two fronts, Qualcomm selling

to ByteDance when Nvidia can't, Arm and

Cerebras, all extends into space. AI

infrastructure here on Earth will play

an important role in the space race.

These are not separate stories. The AI

revolution extends into orbit and to me,

that's a future worth investing in.

Right now, my plan is to wait for

SpaceX's IPO and compare all the space

stocks side by side. Let me know in the

comments if you want me to cover more

space stocks in general or do a deep

dive on SpaceX or Rocket Lab

specifically. 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.