NVDA is undervalued; Rubin scaling could drive quarterly revenue to $1T in ~2 years.

Now, they have some NVIDIA compute on the robot. They have some training compute, but they don't have anything near Tesla's training compute, and they don't have anything near Tesla's volume of training data.

we'll have a SpaceX designed Nvidia rear Ruben NVL72 computer. That's why the thumbnail has a picture of Elon and a picture of Jensen because Jensen is the CEO of Nvidia. Lichy again. Um Jensen is the CEO of Nvidia.

And if you do the math on this, 100,000 satellites with 72 GPUs from Nvidia each, you get to 72 7.2 million GPUs. 100,000 racks is in the ballpark of $10 million each. So that's like, I don't know, trillion dollars.

I think that's a trillion dollars. Um, so that's a big revenue for Nvidia if you're paying attention to that.

I'm invested in Tesla, SpaceX, Tesla, Nvidia, SpaceX, and Google. Those are my four main investments.

You can see that the these are um I think they're racks of six maybe Nvidia GPUs. 72 total. I think there's six of those. So maybe there's like 12 in each rack.

Like Nvidia's revenue last quarter was $96 billion, right? Nvidia is already doing 4x that between Nvidia and anthropic open AI.

Nvidia is the most obvious undervalued stock on the market. They're coming out with Ver Rubin. And I think it's already shipping in small scale. Ver Rubin is going to be shipping at scale probably within about a year.

And at scale is a,000 units a day. That's $10 billion a day. That's like a hundred billion a No, I think that's a trillion dollars a quarter. 10 million times a,000 $10 million times a,000 units times 90 days.

It's like $900 billion a quarter just for Ver Rubin. You still have Blackwell 300. You still have other things that the business does. You're talking about a trillion dollars a quarter.

They did like less than $100 billion last quarter. And you're talking about like two years away they're doing a trillion dollars a quarter. It's insane.