Today, while scrolling Douyin, I saw a news story about Dell founder Michael Dell and his wife, Susan Dell.
Two people put up $6.25 billion to help 25 million American children set up investment accounts.
What does $6.25 billion mean?
Based on the current exchange rate, it’s roughly over 40 billion yuan—around 40-odd billion.
My first reaction wasn’t really, “Having money is great.”
Instead, I suddenly thought of a question:
Does Dell still make money like this?
Because in my impression, Dell is still that old-school tech company that sells computers.
When I was a kid, there was Dell in the school computer lab, and there was Dell in the office.
Later, when people discussed tech companies, the topics slowly shifted to Apple, Nvidia, and Tesla.
Dell seems to have been gone from the most “sexy” stories in the tech industry for a long time.
I just happened to look up its most recent earnings report.
then you find out:
My impression of Dell may be years behind.
In a single quarter, it booked $60.9 billion in AI server orders.
In Dell’s latest quarterly earnings report, there’s a number that’s extremely staggering:
AI server orders: $60.9 billion.
Note—not for the whole year.
it’s a quarter.
In the same period, Dell’s truly recognized AI server revenue was about $16.4 billion.
And by the end of the quarter, it was still holding:
$95 billion in AI server orders awaiting delivery.
In other words:
It’s not that nobody is buying—there are simply too many orders to deliver them all yet.
over the whole quarter, Dell’s revenue hit $47 billion, up 58% year over year.
The company even raised its full-year revenue outlook directly to:
$192 billion.
Only after seeing this did I realize:
Now, Dell is no longer just “a company that sells laptops,” at least not based on current orders.
Nvidia sells chips; Dell sells the entire “AI factory.”
In the past two years, whenever people mentioned AI infrastructure, almost everyone’s first reaction was Nvidia.
That’s of course right.
AI needs GPUs.
But there’s one very realistic problem:
You can’t buy a GPU and just leave it on a desk to run AI.
What do truly large-scale AI systems need?
GPU, CPU, memory, storage, networking equipment, server racks, power supplies, cooling systems…
and finally, it has to combine all these into a data center that can run reliably.
What Nvidia provides is the most core “engine.”
And what Dell is doing now is more like:
Assemble the engine, chassis, circuits, and the whole car, and then deliver it straight to the customer.
For large enterprises, cloud computing companies, AI startups, and government agencies that want to build their own AI compute centers, it’s impossible for every single one to start from scratch and build servers themselves.
This is Dell’s position.
It’s turning Nvidia’s GPUs into AI servers and data center infrastructure that can actually be deployed and used.
So in this wave of AI,
If Nvidia is selling shovels,
then Dell is more like:
an entire mining-equipment setup.
Even more extreme: the $95 billion in orders hasn’t even been delivered yet.
I think the most worth watching in this earnings report isn’t even the $60.9 billion of new orders.
Rather:
$95 billion in backlog AI server orders.
What does that mean?
The customers have already placed orders, but Dell hasn’t yet turned all of those orders into revenue.
From an investment perspective, this number is very interesting.
When a company’s orders suddenly surge, sometimes it’s just short-term frenzy.
But if:
orders keep increasing,
customers keep growing,
delivery capacity keeps up,
and in the end, can actually turn these orders into real profit—
which is a completely different story.
Dell’s number of AI customers increased—from about 5,000 to 6,500—just in one quarter.
over the past 12 months, Dell’s AI server orders have already exceeded:
$130 billion.
This means it’s no longer just a few tech giants “experimenting with AI.”
more and more enterprises have truly started spending money to build AI infrastructure.
A computer company in its 40s has stepped back to the center of the tech wave
Dell was founded in 1984.
Back then, he founded the business by assembling computers in his college dorm.
Later, with the PC era, Dell became one of the world’s largest PC companies.
Later, with the rise of smartphones and cloud computing, the PC industry gradually became less sexy.
Many people’s impression of Dell still stops at that era.
But what’s interesting about the tech industry is exactly this.
A company can lie dormant for many years,
and then the next technological revolution suddenly appeared,
The capabilities it accumulated over decades have once again become valuable.
AI happens to require a large number of servers.
And servers are precisely one of the things Dell has been most familiar with for decades.
It doesn’t even need to completely reinvent its business model.
Simply take what it already does well and amplify it again in the AI era.
Now the whole market is looking for the so-called:
“The next Nvidia.”
But the more I think about it, the more I feel the question itself may be asked the wrong way.
With an industry this big, there can’t be only one winner.
Some sell GPUs.
Some sell memory.
Some build data centers.
Someone provides power.
Some people are also responsible for taking hundreds of thousands of GPUs and actually assembling them into AI infrastructure that can run.
And right now, Dell is standing in that position.
The company has already raised its AI server revenue forecast for the current fiscal year to:
$74 billion.
So next I’m looking $DELLB —what I truly want to observe isn’t:
“Is there still demand for AI?”
At minimum, based on current orders, the answer to this question is already pretty clear.
The real question is:
Can Dell turn the $95 billion in AI orders it holds—one deal at a time—into revenue and profit?
If possible,
So in this AI gold rush,
The most profitable might not be only the chip sellers.
those companies responsible for turning chips into real “AI factories,”
It’s also something that deserves to be rediscovered.
Personal viewpoint · For record and sharing only
Stay curious and make independent judgments.