Have you ever watched a stock you love get absolutely crushed for reasons that feel more like panic than reality? That’s exactly what’s happening right now with one of the purest AI infrastructure plays on the market.
I’m talking about CoreWeave – the Nvidia-backed “neocloud” that rents out massive clusters of GPUs to anyone training the next generation of artificial intelligence models. A month ago the stock was flying. Today? Down 37% in a single quarter. The headlines are brutal, the Reddit threads are apocalyptic, and half of Wall Street seems convinced the AI party is already over.
But here’s the thing: sometimes the best time to buy is exactly when everyone else is running for the exits.
One Veteran Investor Just Said “Buy This Dip” – Loudly
Paul Meeks has been investing in technology since 1992. He’s managed billions, survived multiple tech wrecks, and he just initiated coverage on CoreWeave with a buy rating and a $100 price target. That’s roughly 17% above where it closed yesterday – and he thinks it will eventually retest its October high around $153.
His basic thesis is simple but powerful: yes, the AI trade will cool eventually. Of course there will be pullbacks. But right now the market is dramatically overreacting to what amounts to a temporary construction delay at one data center out of 41.
“The stock significantly overstates the risk that it will not deliver its $56 billion multi-year revenue backlog.”
Paul Meeks, Freedom Capital Markets
What Exactly Is a “Neocloud” Anyway?
Think of traditional cloud providers – Amazon Web Services, Microsoft Azure, Google Cloud – as the department stores of computing. They sell everything to everyone.
Neoclouds like CoreWeave are more like specialty performance shops. They focus almost entirely on delivering the absolute highest-density GPU compute possible, optimized for training and running giant AI models. No bloated general-purpose services, no distractions. Just raw, screaming-fast AI horsepower.
And right now, demand for that horsepower is still insane.
The Backlog Everyone Is Ignoring
CoreWeave currently sits on $56 billion of contracted revenue stretching out multiple years. That’s not a hope or a forecast – that’s signed contracts with some of the biggest, richest companies on Earth.
- Massive hyperscalers who view AI leadership as life-or-death
- Companies with essentially unlimited balance sheets
- Customers who pre-pay or commit capital upfront to secure capacity
These aren’t flaky startups that might disappear tomorrow. These are the giants who have already decided that falling behind in AI is not an option.
Nvidia: The Ultimate Backstop
Let’s not forget who owns 6.56% of CoreWeave. Nvidia isn’t just a supplier – they’re an investor with skin in the game worth over $2 billion today. They were there before the IPO, they bought more shares during the IPO, and they have every incentive to see CoreWeave succeed.
In fact, Meeks points out that if – and it’s a massive if – some customers ever tried to walk away from their commitments, Nvidia alone could absorb the excess capacity for seven years. That’s how deep this relationship runs.
The Nebius Comparison Is Almost Comical
There’s another public neocloud out there called Nebius. Great company, up 250% this year, everyone loves it.
Here’s the punchline: Wall Street currently values Nebius at 7.3 times 2026 revenue. CoreWeave? Just 2.8 times using conservative estimates.
Yet Meeks projects CoreWeave’s 2026 revenue will be nearly five times larger than Nebius. Same business model, same customers, same explosive demand – but trading at less than half the multiple.
If that doesn’t scream mispricing, I don’t know what does.
Why the Sell-Off Feels So Overdone
Look, nobody is saying CoreWeave is perfect. They cut guidance because one data center build slipped. Construction delays happen – especially when you’re trying to wire up hundreds of thousands of the hottest GPUs on the planet while the entire industry is scrambling for power and transformers.
But turning a single scheduling hiccup into “the AI bubble is bursting” feels like the market looking for any excuse to take profits after an enormous run.
In my experience, these are exactly the moments that separate long-term winners from the crowd. When fear is highest, when the narrative flips overnight, when perfectly good companies get thrown out with the bathwater – that’s often where the real money is made.
Where Wall Street Stands Today
Analyst price targets on CoreWeave are all over the place – from $36 on the low end to $234 on the high end. The consensus sits around $131, which would mean more than 50% upside from current levels.
Meeks’ $100 call actually sits toward the conservative side of that range, which tells you something about how far sentiment has swung.
The Bigger Picture for AI Infrastructure
Step back for a second and think about where we actually are in the AI buildout cycle.
We’re still in the very early innings. Most companies are barely past the “let’s throw some GPUs at this and see what happens” stage. The real production-scale deployments – the ones that will require orders of magnitude more compute – are still ahead of us.
And every single one of those deployments is going to need exactly what CoreWeave sells: dense, efficient, purpose-built AI cloud capacity.
The hyperscalers can’t build it fast enough themselves. The startups can’t afford to buy the hardware outright. That leaves the neoclouds perfectly positioned in the middle.
Risks You Should Actually Worry About
To be clear, this isn’t a risk-free investment. Debt levels are high – they have to be when you’re financing tens of billions in GPUs. Competition is increasing. Power availability is becoming a real bottleneck industry-wide.
But these are known risks. They’re the same risks every player in this space faces. The question is whether the current price already reflects those risks or whether it’s gone too far into panic territory.
Right now, I’d argue it’s the latter.
The Bottom Line
CoreWeave isn’t going to zero. The $56 billion backlog isn’t suddenly disappearing. Nvidia isn’t walking away. The secular demand for AI compute isn’t reversing.
What we have is a temporary sentiment low on a company that sits at the absolute center of one of the most important technology shifts of our lifetime.
Seasons change. Narratives flip. But companies that control critical infrastructure during major platform shifts tend to do just fine over time.
Sometimes the best opportunities come wrapped in fear. This might be one of those times.
Disclosure: The author may hold positions in companies discussed. This is not investment advice – always do your own research.
