4H·

What does Nebius do? Simply explained!

attachment

What does $NBIS (-1,1 %) ? And what do they do differently from the big players like Google $GOOGL (+2,21 %) , Amazon $AMZN (+0,97 %) and new companies like $CRWV (+0,94 %) and $IREN (-0,2 %)?


Many people simply find it difficult to understand the business model. That's why I'll try to explain it in a way that even someone who has nothing to do with technology will understand.


Imagine the computer world as a huge playground where companies build super-fast machines to develop smart robots or cool AI games, for example an AI that paints pictures or can chat with you like a friend. $NBIS (-1,1 %) is one of these companies, but it is special. Because they build their own slides and swings, i.e. their own infrastructure for AIs, and don't just use old, boring ones from others.


The fact that they build their own playground is VERY IMPORTANT and I'll explain why.


Once they've built this playground, they rent out these super-powerful computer parts - called GPUs (think of them as magical brains for AIs) - to people who want to use them to develop cool AI stuff.


The competition is other "playground builders", but big ones like $AMZN (+0,97 %) (AWS) or $GOOG (+2,27 %) (GCP) who build playgrounds for all kinds of games (not just AI), and smaller ones like $CRWV (+0,94 %) or $IREN (-0,2 %) who also focus on AI but have a different approach.


Like I said: $NBIS (-1,1 %) builds everything themselves (a kind of "do-it-yourself" kit), unlike most others who buy pre-made parts from third parties. Nebius develops its own computers and software, specifically for AI. As a result, their machines run faster, consume less energy and cost less to rent. This is why Nebius' prices are 30-40% cheaper than AWS or GCP and the reason for these lower prices is important, as explained above. It's like building your own fast bike instead of buying a slow one from the store.


This also allows them to get the latest and greatest GPUs from Nvidia faster than others. Nvidia $NVDA (-1,21 %) is also a partner (i.e. a buddy).


The big competitors like AWS or Google don't just do AI, but also a hundred other things, like storing your vacation photos or running websites for cupcake stores.


$NBIS (-1,1 %) says: "No thanks, we only do AI!" and that's why their tools are particularly well suited to AI tasks, such as training or efficiently running smart bots.


It's like a store that only sells ice cream and therefore has the tastiest ice cream. Would you rather go there or to a supermarket that sells a thousand other things?


Or imagine a family doctor who can do a little bit of everything compared to a neurosurgeon who specializes in brain surgery. Who do you go to when you need brain surgery? Exactly.


Then you ask: What about $CRWV (+0,94 %) and $IREN (-0,2 %) ?


Imagine building your own playground, which costs a little more - but you also sell cupcakes, juices and summer drinks on the weekends. So you have a lot of savings and additional revenue streams to rely on if your playground ever needs more money. $NBIS has over $1 billion in cash and three other growing businesses, as well as 28% in a fourth company to fund their AI computing. This ability to raise capital quickly and without dilution is extremely important.


They build themselves.

They build efficiently.

They build faster.

They build cheaper.

They charge less money.

And they have enough money to build even more.

14
3 Commentaires

image de profil
I recently started a position and get more excited about their potential the more I read up about them. The market also expects enormous growth so anything less than phenomonal is likely to get punished after their next earnings.

Btw you meant $IREN instead of $IRE
image de profil
@Nova_portfolio I also opened a position today. And thanks for the reminder, I fixed it with $IREN !
image de profil
I don't see many reasons why you should invest in small cloud providers. Almost all of them make losses, did something else before and have huge problems. In contrast to Amazon and co, they have the problem that you can't always just add the latest chips somewhere. If you do it, you basically write off your own chips again and again. There is simply a lack of scale. In addition, the basic costs are simply higher at 10 billion than at 1000 billion. They can hardly do research and the like in a meaningful way
Participez à la conversation