Nvidia's Trillion-Dollar $1B+ AI Startup Web  ## Document Notes ## Highlights - **NVIDIA Became The AI Ecosystem's Core** - NVIDIA transformed from a chip seller into a strategic investor fueling the AI ecosystem by investing across many top AI firms. - Jaeden Schafer argues this expands NVIDIA's influence and creates a feedback loop where invested companies buy NVIDIA hardware. Transcript: Jaeden Schafer I think it's safe to say that no company has ridden the AI wave quite like NVIDIA. Ever since ChatGPT first came out, which is, of course, over three years ago now, we've seen, you know, NVIDIA has seen explosive growth from all of their AI products that followed and All of the AI products that followed relying heavily on NVIDIA's chips. And so, of course, the revenue, the profits, their cash reserves, everything surged, including their stock price, which turned them into roughly a $4.6 trillion company. So right now, they are the world's dominant supplier of high-performance GPUs, and they have used their financial firepower to dramatically expand their investing, especially In different AI companies and infrastructure deals. Because of course, if they put money into AI companies, all of the infrastructure required to run it is going to get, you know, they're going to be buying NVIDIA GPUs and pay money back To NVIDIA. So just last year alone, NVIDIA did 67 venture deals. And that is more than their 57 investments that they made the year before in 2024. So all of those numbers don't include all of the deals that were completed through NVIDIA's former corporation venture arm, Nventures, which also ramped up significantly. So according to PitchBook, Nventures completed 30 deals last year compared to just one in 2024. NVIDIA also said that they are investing strategically and focused on growing the broader AI ecosystem by backing startups that they believe can become true market makers. And of course, we see that this is directly beneficial to them. ([Time 0:01:16](https://share.snipd.com/snip/05187be9-27cf-4bf4-a7f4-5ad5e18bbd6d)) - **Investments Drive Hardware Demand** - NVIDIA's investments often pair capital with expected hardware purchases, aligning startup growth with GPU sales. - Jaeden Schafer frames this as strategic round-tripping that benefits NVIDIA financially and operationally. Transcript: Jaeden Schafer So I want to go over a list of startups that raised rounds larger than $100 million since 2023 with NVIDIA named as an investor. We'll order them from largest to smallest round sizes. And I think just all of these combined should show just how far NVIDIA's influence now extends beyond just, of course, selling chips, but they are getting ownership shares in every Single major AI company at this point. So the first one, of course, is OpenAI, the big company behind ChatGPT that we all use. And NVIDIA's put their, they've essentially first invested in them in October 2024. And they invested apparently, reportedly $100 million in their $6.6 billion round that back then valued them at $157 billion. So NVIDIA didn't actually appear in OpenAI's $40 billion round that closed in March last year, but they later announced a strategic partnership that involved another $100 billion In infrastructure investments. Although NVIDIA said that, you know, this deal was not guaranteed to close as it was outlined. And so to take it with a grain of salt, which I actually appreciate that clarity from Nvidia, because I feel like we see a lot of these deals where they're like leaked or rumored. ([Time 0:02:45](https://share.snipd.com/snip/3197f27f-aaf3-4131-ae98-3b4cf2a7c8b4)) - **Anthropic Deal Tied To Hardware Commitments** - Jaeden describes NVIDIA's November investment in Anthropic as part of a strategic round that included Microsoft. - He highlights Anthropic's commitment to spend tens of billions on NVIDIA hardware as part of the deal. Transcript: Jaeden Schafer As part of that arrangement, Anthropic agreed to spend tens of billions of dollars on cloud compute and NVIDIA hardware. So again, it's a direct, we give you money, you buy our hardware. Cursor is the most popular AI code editor out there. It helps developers write, refactor, understand code faster. And NVIDIA made a first strategic investment in November as part of a $2.3 billion Series D that valued them at $29.3 billion. NVIDIA's kind of doing that is kind of their official entry as a shareholder after years of just being an enterprise customer. ([Time 0:04:51](https://share.snipd.com/snip/f1e72bbb-0565-4b0e-b483-78a88475c3aa)) - **NVIDIA Funds Rivals Despite Pressure** - Jaeden recounts NVIDIA investing despite industry pressure against funding rivals, noting NVIDIA's independence. - He points out NVIDIA's scale makes it able to fund competitors like XAI without deferring to other firms' wishes. Transcript: Jaeden Schafer They have Grok, which is incorporated into Tesla as kind of the voice assistant and also incorporated into X or formerly Twitter, the platform where it can break down stuff. There's like a lot of controversy with it. Despite OpenAI previously encouraging some of their investors to avoid funding rivals, which of course is just ridiculous. And you can't tell NVIDIA who to not fund because NVIDIA is, you know, NVIDIA is the market maker. It's a massive $4 trillion company. Despite all that, NVIDIA participated in OpenAI's $6 billion round last year, and they were expected to invest up to $2 billion more as part of OpenAI's planned $20 billion raise. They had a deal that was kind of structured in part to fund additional NVIDIA hardware purchases. ([Time 0:05:33](https://share.snipd.com/snip/66d65ae3-2397-4973-8a2c-60fa017ee41c)) - **Global Diversification Of Investments** - NVIDIA invests internationally (e.g., Mistral in France) to avoid being US-only and to access global AI markets. - Jaeden Schafer sees repeat investments as NVIDIA diversifying geographic influence across AI companies. Transcript: Jaeden Schafer They have this open source, and they also have a commercial large language model that's designed to compete with US and Chinese labs. It's not as good as OpenAI, if I'm speaking frankly. And by the way, every model I've mentioned so far, you can go try on AIbox.ai, my platform. We have a playground where you can test out every model against each other and show the results side by side. But Mistral is a big company. They're doing great and they're really focusing on enterprise adoption. They're focusing on a lot of the needs of Europe. NVIDIA invested for the third time when Mistral raised their 1.7 billion euro round. It was their Series C. They did that in September last year. That put them at about a $13.5 billion valuation. And NVIDIA kept coming back to invest in Mistral. ([Time 0:06:24](https://share.snipd.com/snip/cf287fbf-e60c-479c-ba4b-4dca643544a0)) - **Betting On Open Source And Talent** - NVIDIA backs open-source and early-stage research plays like Reflection AI and Thinking Machine Labs to capture future tech leaders. - Jaeden Schafer highlights massive valuations raised pre-product as evidence of clout-driven funding. Transcript: Jaeden Schafer And I think when you say lower cost, I would also argue that means like less high quality, but it's open source, which is amazing. So in October, NVIDIA was one of the largest investors in their $2 billion round, which valued, it's only a one year old startup, but it valued them at $8 billion. And they're positioning themselves as a US counterweight to China's DeepSeek. So they're like, look, we're going to, obviously DeepSeek is kind of this open source model out of China that's got a lot of attention, but it's not as good as OpenAI, but it's open source We want to do that same thing but inside of America. The next one is Thinking Machine Labs so this is a new AI research company it was founded by Miriam Ratti who is OpenAI's former CTO and NVIDIA joined a long list of investors that gave Them about two billion dollars in a seed round they did in July last year. They were valued at $12 billion. They haven't even released a product. It's just Miriam Moratti, the former OpenAI CTO. She's got enough clout. ([Time 0:07:18](https://share.snipd.com/snip/cb6d1e5a-27e5-44d6-aa07-723c8b74dcc5)) - **Funding The Data Center Buildout** - NVIDIA also invests heavily in AI infrastructure firms building data centers for large projects like OpenAI's Stargate. - Jaeden Schafer argues this doubles-down on NVIDIA's role in both compute supply and the physical infrastructure buildout. Transcript: Jaeden Schafer But for all intents and purposes, it basically got acqui-hired by Microsoft. Crusoe. Oh, and by the way, when all of that happened, I think all the investors didn't get payback. So NVIDIA would have made money back on that deal. Okay, Crusoe builds massive AI-focused data centers and also energy infrastructure. And NVIDIA joined their $1.4 billion Series E in October last year, which gave them a $10 billion valuation as they were expanding their data center campuses for OpenAI's Stargate Projects. So obviously, they're working directly with one of these big projects. Nscale builds AI optimized data centers, but they do it over in Europe. NVIDIA participated in their $1.1 billion round. They also joined a $340 million safe financing, which was tied to the company's role in OpenAI Stargate infrastructure. So it's kind of interesting because OpenAI has got this big deal, but in this big Stargate program, we're going to spend $500 billion. But in order to do it, they need like a lot of compute and infrastructure. So they got all these infrastructure companies who go raise money to build out the infrastructure for open AI. I know it's just it's crazy the amount of money. So I mean, in all you know, to give NVIDIA a lot of clout and like credit, they are making an insane amount of money, but they are dumping it right back into the whole industry. ([Time 0:09:15](https://share.snipd.com/snip/61c5e5eb-ec0f-4b6f-9762-538fcaaee4ac)) - **Broad Portfolio Across The Stack** - NVIDIA's portfolio spans model builders, tooling, and data-labeling firms like Scale AI, embedding them across the stack. - Jaeden Schafer notes this creates broad influence and potential control across model development pipelines. Transcript: Jaeden Schafer Scale AI gives data labeling infrastructure services, which are used to train LLMs. So you can imagine like, I've looked at kind of what ScaleAI does. They'll go find a whole bunch of lawyers. They'll give a whole bunch of legal stuff and have the lawyers label the data and give them like kind of this customized data labeling, which you don't get anywhere else. They employ an insane amount of people. I want to say, I don't want to misquote the number, but they employ a lot of people and they're highly specialized like lawyers, doctors, and they're going and labeling specific legal Or medical data and explaining what it's used for. And then they go around and sell this data back to AI companies. Sometimes it's contracted ahead of time. In any case, NVIDIA joined their $1 billion round that they did in 2024. And that was before Meta later bought 49% of the company. And they also hired Meta's or Scales CEO. And it's basically kind of like another one of these Aqua hires, but they only bought like half the company. Now they can basically control what happens to it. ([Time 0:11:20](https://share.snipd.com/snip/6e27cdcd-d142-405a-926c-ea4cfee785d5)) - **Equity Plus Hardware Creates A Flywheel** - Jaeden interprets NVIDIA's strategy as strategic and mutually reinforcing rather than merely speculative or round-tripping. - He believes equity stakes plus hardware demand position NVIDIA to keep financing and benefiting from AI growth. Transcript: Jaeden Schafer I believe that this is a great strategy for NVIDIA because they're also getting equity in all of these companies. And so NVIDIA's portfolio is growing and is absolutely impressive. Essentially, they're like financing the entire industry while their stock price goes up so they can continue financing the entire industry and their stock can keep going up. It's kind of crazy. A lot of people call it a bubble or cyclical or, you know, round tripping the money. But I think at the end of the day, it's very strategic and a good move on NVIDIA's part because at the end of the day, now they own a massive chunk of tons of the top companies in this massively Growing industry. ([Time 0:12:59](https://share.snipd.com/snip/c9e1d9bd-b9ab-489f-bd71-1366453bd944))