NVDA收购Hugging Face是一项防御性的长期战略，旨在维持生态系统主导地位；该收购不会立即产生现金资产。

Nvidia has just changed the AI market with its recent acquisition of the "Hugging Face" platform. There are also other updates announced by Nvidia that I think every AI investor should know about.

As I mentioned, I really want to focus on Nvidia and this investment in " Hugging Face". Now, I think this is actually a very big and very important thing. This highlights how far ahead Nvidia is in this field. In my opinion, it also shows how concerned Nvidia is to some extent about someone coming along and taking over its position in general.

So, Nvidia agreed to acquire " Hugging Face" for $12.9 billion.

Because the first thing Nvidia wanted to make sure of was informing investors and the world, especially developers, that " HuggingFace" would remain an open platform for all AI systems.

Well, firstly, Nvidia is already a major contributor to open-source solutions. They've talked about it in the past, and even shown explanatory slides, haven't they ? They have released more than 500 models on the " Hugging Face" platform and more than 250 open datasets. They have built many different solutions here, and shared them with the world in open-ended terms.

The way I see it is that Nvidia has always been afraid or worried about becoming obsolete in one way or another, hasn't it? Some perfect examples are that Nvidia thought that cloud service providers would no longer need Nvidia . So, what did Nvidia do? They began innovating at rapid paces. They started investing in new players in the field of cloud computing. They started investing in industries within the artificial intelligence system to make Nvidia's position extremely broad , didn't they? There is no longer a single customer on whom Nvidia's future depends entirely. Nvidia made sure there were plenty of customers and gamers to make that happen.

I really think Nvidia suffers from a bit of paranoia. What will happen if we continue in the future to see open AI models led by Chinese players ? Will these Chinese players eventually start improving those models mainly on Chinese AI chips or something outside of the Nvidia system, thus reducing the demand for their products, right?

So, I now think that Nvidia has this suspicion that there could be a world where only a select group of companies owns the best AI models , maybe two or three companies.

It will be something like Nvidia, where you will continue to progress with it. It will continue to innovate within the chip ecosystem, and that is where Nvidia succeeds. I think Nvidia is really paranoid that in the world of the future, if there are a limited number of AI companies , those companies will be able to restrict or control distribution as they see fit. This would be bad for Nvidia's business, because those companies could make their own custom chips. Instead, I feel that Nvidia is making sure that there is this huge open-source market for weights . There won't be one or two winners, and this will allow Nvidia products to deliver very good performance.

Interestingly, I feel that this acquisition may face some kind of objection from other players. How can Nvidia own the open- source ecosystem or the open world, right? This doesn't seem logical. So, I can expect some objections here, but it's great to see other CEOs agreeing to this acquisition. Yes, I mean I think the acquisition is good, at $13 billion. I don't believe this acquisition will bring Nvidia any cash assets in the near term or anytime soon . It's a long- term game to ensure the ecosystem stays within Nvidia's reach, because there will be more than two players in this race, and there will be a lot of open-weight models as I'll explain shortly, and I think that's the real win. It is a market where your promise of a computing world is not limited to a select few. Rather, he promised a computing world that contains many different models, and all of these different models will work on the Nvidia ecosystem.

Nvidia announced two impressive things today. First, they announced RTX PCs, which are laptops expected to get closer to the world of artificial intelligence. For those unfamiliar, it appears that Nvidia feels it is missing a major opportunity in AI computing for the consumer market.

Therefore, Nvidia released the DGX Spark device, but it runs on Linux . It is not something that most ordinary consumers use. I have a DGX Spark here, but they're now releasing the RTX Spark, which will be a Windows PC, and it will be released in October. We finally got a date in October. We still don't know the prices yet. But this shows the huge opportunity that Nvidia realizes it is missing, while Apple captures all the market share in this area. I am absolutely certain that they will achieve great sales here. I am very likely to buy one of these devices, and I think they are an excellent match for these types of open- source models.

However, there will be a fair number of open- source models that will run locally in products like these . Therefore, Nvidia will be able to take advantage of this . I'm very curious about the pricing of these products due to the current oversupply in the memory market, but at least we now have an arrival date in October. They also discuss how to streamline native AI support for Nvidia's graphics processing units.

Now, there is something else happening here with regard to local artificial intelligence, as Nvidia has created something called "Bear," which is Nvidia's personal AI mentor. What happens here is that Nvidia allows you to connect to other Nvidia products, whether it's a laptop, a desktop with an Nvidia graphics processing unit, another DGX Park device, or another laptop. It's like setting up your own small data center, isn't it? It's like saying, "Hello, I have all this computing power. How can I break this task down using all this power I have?" "So, perhaps the task can be completed earlier or faster." Therefore, local artificial intelligence has become exciting in my opinion . I truly believe that we are living in this era of native artificial intelligence. And I still believe that we are strongly in the age of cloud-based artificial intelligence. This will continue for a long time. But local artificial intelligence is becoming increasingly popular, especially as we see Nvidia investing in many of these areas. So, I thought this was very cool. The Nvidia Bear Beta application connects AI applications and workflows to a single local endpoint to route inference across Nvidia DGX Park, Windows systems with RTX, other types of GPUs, and Mac OS . This helps you maximize local computing while maintaining the privacy of command files and the proxy context.

Oh, yes, there were reports that Nvidia was considering "Thinking Machines" as the company held talks to raise $1 billion at a valuation of $40 billion. This is a startup company in the field of artificial intelligence. I don't remember if they will make it fully open source. I think they might do that. But regardless, Nvidia is looking to invest money in this, and I think we will continue to see a lot of innovation in the world of artificial intelligence.