英伟达在半导体领域的主导地位已定；所有产品都将售出，这一趋势将持续3-4年。

You don't need to waste any more brain cells worried about Nvidia,

AWS and Nvidia striking a strategic partnership for $2 million 2 million additional GPUs. It's going to cost way more than $2 million. 2 million Nvidia GPUs is billions of dollars. This is on top of the 1 million GPUs that Amazon agreed to buy starting in 2026.

Moving on to Nvidia. Started the week at 212 and on a week where it announced earnings, shares exploded after earnings, but then drifted sideways and down, but finished the week higher by 2.3%.

The company announced those Q2 earnings revenues came in at $96 billion. They are guiding for 105 or they had 105% revenue growth in the quarter. More specifically, the company guided to their fiscal 2027 including the third quarter, but really all next year. They're expecting growth next year over 70%. They actually said, I think they alluded to it, that it'd be actually triple digits if they didn't have so much supply constraints.

When you look at the financials, they're incredible. The company's net income after everybody gets paid was $60 billion in a quarter. There are companies that don't do 60 bill there are fortune 50 companies that do do $60 billion of revenue in 10 years. This company is doing $60 billion in net income. It is absolutely extraordinary.

Now the company also preparing an acquisition of hugging face which is an AI repository. It's often the place where you go to launch the model. We've actually seen two critical acquisitions over the last week or so. Stripe bought open router. Nvidia buying hugging face. What this is showing me is these companies are preparing for a world where one model doesn't win. where you are going to have dozens if not hundreds of models that win. If you are of that worldview that it's not going to be just anthropic and just anthropic [clears throat] and open AI. If you are the worldview that there are going to be dozens if not hundreds of AI models that people are going to be using on a daily basis, then hugging face and open router are two key acquisitions. I think it's interesting that you have Stripe, which is an incredible company, and then obviously Nvidia goes without saying, it might be the most incredible company, making acquisitions in that space. It's telling you where they think the industry is heading. They're making acquisitions here, and I think it's going to end up being a smart one.

Now, the company is pausing, though. It's revenue sharing deals with AI cloud partners. This is where they made some venture investments in companies where they offered credits at AI cloud providers in exchange for a share of revenues. Some people citing potentially regulatory concerns here, although I I don't think the Trump administration is worried about that. So, this must be something else. Maybe there was a lack of interest. Maybe there was a lack of demand or there was a lack of resources over at Nvidia to pull this off.

Nvidia is guiding forward a year. Anything that these companies make is going to get sold. Over the past 3 years, we had to discuss if Nvidia's revenue was going to continue at this trajectory. We had to discuss if their technology was going to get adopted after each iteration and each price increase. We had to discuss if OpenAI and Anthropic and Amazon and the hyperscalers were going to adopt their technology. Folks, all of those discussions are done and over. There is nothing left to discuss here. This would be like discussing the difference between Google search engine and Yahoo search engine or whether the iPhone is still going to be around in three or four years.

At this stage, guys, there is nothing to discuss in the semiconductor industry. Whatever Nvidia makes is going to get sold. Now, don't mistake that with, oh, we need to run out and sell our semiconductor stocks. That's how a kindergarten would think. What I am saying is there is nothing left to discuss in semiconductors. There are things to discuss in AI. There's things to discuss in physical AI. There's a lot to discuss around how AI is going to be implemented across businesses, organizations, governments, and society. But there is literally nothing left to discuss at the semiconductor layer. You have three or four dominant companies at the top. You have one to two manufacturers that dominate. You have one to two memory companies or three memory companies that dominate. And that's it. That's all you need to know. And the trend is going to continue to play out for the next three or four years. But trying to discuss the new chips is like discussing the new iPhone. It's going to come out. It's going to sell. And that's all you need to know.

What you have to understand is this SpaceX is going to launch these into the sky. This is going to be Nvidia that is going to develop this. So you are doubting Nvidia developing this. The other thing you have to understand about space data centers. First of all they already exist. Number two, you subtract two of the hardest and most costly aspects of a GPU or a data center. The power is more or less solved and the cooling is 100% solved. You don't have to cool these things in space and the sun shines essentially 24 hours a day at the right orbit in space. So you solve those two things and this is not that complicated. Now, getting the data to and from Earth to the data center, that's probably a little bit harder as well, but something that is probably can be solved.

Nvidia earnings weren't enough to power it higher.

Nvidia made a beautiful move after earnings and then gave most of it back. We talked about how they were up only just south of 3% this week. You retested the highs. You didn't create a new higher high on the chart, but all your moving averages remain positive.