@AdityaInvests90 All around. It’s the better cloud, has far more active and contracted compute, and better debt terms once you weigh the equity dilution.
$CRWV > $NBIS
The “CoreWeave has more debt, therefore bad” argument is a strange one from people who don’t seem to understand how GPU infra gets financed. You fund GPUs with debt!
That is largely why CoreWeave has more capacity, a much larger backlog, and stronger pricing power.
CoreWeave has better terms for actual GPU backed project finance. Nebius has cheaper headline financing because it leans more on converts, equity linked capital, Nvidia warrants, etc. Nebius effectively shifts the cost into dilution, whereas CoreWeave (smartly) finances unlimited demand because they are the best at what they do and have unlimited demand for their product.
So everyone parroting that Nebius is better than CoreWeave is unknowingly buying the right to be diluted. I guess ignorance is bliss?
You do realize one funds GPUs with debt, right? 🙄
You also realize there is a reason CoreWeave has much more GPUs, and is able to charge higher prices, right?
CoreWeave has better debt terms for actual GPU backed project finance.
Nebius has cheaper headline financing because it relies more on converts, equity linked capital, Nvidia warrants, etc. It’s cheaper cash interest that literally shifts cost into dilution for folks like yourself.
The greatest risk to the compute buildout is the reality that it takes time to build data centers and a lot of data centers will realistically be delayed.
This also means that as compute demand scales, and supply increases (but delays ensue), the price per GPU hour increases for available supply.
Compute is a great business.
Wondering why tech sold off today?
Burry said “bubble” 3x fast into a mirror.
Fortunately, the genie only grants him 1 wish per quarter, and the spell usually wears off quickly, once the market remembers AI is a supercycle.
Bet Burry is feeling really good today.
Shame he’ll look like a fool again in a week. And that he still looks like a fool of you zooming out a mere week.
Great buying opportunity for those who are not emotional traders.
Just modelled $CRWV earnings through 2030 based on +5GW active capacity target.
It gets gradually derisked as interest expense declines from almost 25% of revenue this year to below 7% in 2030.
In 2030, Debt/EBT ratio declines below 9, which is still high but manageable.
So, if the backlog converts, there is nothing that justifies its discount relative to $NBIS. If AI capex slows down and backlogs don’t convert, nothing is saving $NBIS as well.
Customer concentration argument is also irrelevant because bulk of projected $NBIS revenues are also from just two companies, $MSFT and $META.
In short, $CRWV being valued less than $NBIS is just ridiculous.
A year from now people will look back and say $CRWV was an obvious opportunity.