Nvidia's Ambitious Investment Landscape: 71 Deals, Billions of Dollars, Betting on the Entire AI Era

When we talk about the AI revolution, there is one name that cannot be bypassed - NVIDIA.

This former gaming graphics card manufacturer has now become the hottest giant in the AI era. Since the emergence of ChatGPT, NVIDIA's stock price has skyrocketed, and its market value has exceeded $4.5 trillion, firmly sitting on the throne in the global high-performance GPU field. But what's even more astonishing is that this company is weaving an investment network covering the entire AI industry at an amazing speed.

According to PitchBook data, as of now in 2025, NVIDIA has participated in 50 venture capital transactions, exceeding the 48 transactions in the whole of 2024. And its official corporate venture capital fund, NVentures, is going all out: it participated in 21 transactions in 2025. NVIDIA is building a huge AI ecological empire - from large model R & D to computing power infrastructure, from autonomous driving to medical AI, from nuclear fusion energy to optical chips. NVIDIA can be seen in almost every cutting - edge field related to AI.

This is a high - stakes gamble worth tens of billions of dollars. Today, Silicon Rabbit will dissect this investment list to see how NVIDIA uses capital to lock in the next decade of AI.

01

In the world of AI, large models are the "brains". Whoever masters the most powerful brain will have the say in AI. NVIDIA understands this well, so we are seeing an unprecedented multi - pronged bet.

The story starts in October 2024. At that time, OpenAI completed a $6.6 billion financing round, and its valuation reached $157 billion. In this financing round, NVIDIA made its first move, investing $100 million. This was already a significant amount at that time.

But no one expected that less than a year later, NVIDIA announced an even more crazy plan: in the form of a strategic cooperation, it will invest a cumulative $100 billion in OpenAI to jointly build large - scale AI infrastructure.

What does $100 billion mean? It means that NVIDIA is deeply tying itself to the OpenAI battleship. Every iteration, every new function, and every training of ChatGPT will consume a huge amount of GPU computing power. And almost all of this computing power comes from NVIDIA.

This is a win - win cooperation: OpenAI needs NVIDIA's computing power support, and NVIDIA needs OpenAI to prove the commercial value of AI. When the whole world is using ChatGPT, NVIDIA's GPUs become the "money - printing machines" of this era.

But NVIDIA is not satisfied with just betting on OpenAI. In fact, its investment strategy is more like "casting a wide net" - no matter which technology route explodes, NVIDIA will be involved.

The story of xAI is the most dramatic. It is an AI company founded by Elon Musk, directly competing with OpenAI. In December 2024, NVIDIA participated in xAI's $6 billion financing round. Even more exaggerated, in xAI's subsequent planned $20 billion financing, NVIDIA is prepared to invest up to $2 billion in the equity part to help xAI purchase more NVIDIA devices.

There is a subtle detail here: OpenAI clearly opposed NVIDIA's investment in xAI because the two companies are direct competitors. But NVIDIA ignored this opposition. In NVIDIA's view, the diversification of technology routes is more important than taking sides. Who will win, Musk or Altman? NVIDIA's answer is: I want them all.

Looking across the Atlantic Ocean, Mistral AI is the rising star of European large models. This company, headquartered in France, completed its Series C financing in September 2025, raising a whopping 1.7 billion euros (about $2 billion), and its post - investment valuation reached 11.7 billion euros (about $13.5 billion). It is worth noting that this is already NVIDIA's third investment in Mistral AI.

Why invest in the same company three times? The answer is simple: Mistral AI represents Europe's autonomy in the AI field. In a situation where US large models dominate the world, Europe urgently needs its own "ChatGPT". And NVIDIA is betting on this geopolitical variable.

In October 2025, a younger name came into view - Reflection AI. This company, which has only been established for 1 year, received a $2 billion financing led by NVIDIA, and its valuation directly reached $8 billion. Its positioning is clear: to develop an open - source large language model to compete with DeepSeek from China.

Against the backdrop of the Sino - US technological competition, NVIDIA leading the investment in Reflection AI sends a profound signal: it wants to ensure that the US does not lag behind China in the field of open - source large models. This is not only a commercial investment but also a battle for technological sovereignty.

There is also a name that cannot be ignored - Thinking Machines Lab. The founder of this company is Mira Murati, the former chief technology officer of OpenAI. In July 2025, it completed a $2 billion seed - round financing, and its valuation reached $12 billion. Getting $2 billion in a seed - round financing and having a valuation of $12 billion is almost unprecedented in the history of venture capital.

Why? Because Mira Murati was directly involved in the R & D of GPT - 3, GPT - 4, and ChatGPT during her time at OpenAI. Her technical ability and industry connections make investors believe that she can create the next world - changing large model. And NVIDIA, of course, will not miss this feast.

From OpenAI to xAI, from Mistral AI to Reflection AI, from Thinking Machines Lab to many other large model companies, NVIDIA's logic is clear: it doesn't bet on a single winner but on the entire track. No matter who eventually emerges on top, they will need to use NVIDIA's GPUs to train their models. This is NVIDIA's strategy in the large model battlefield: to make itself the "infrastructure" that all players rely on.

02

If large models are the "brains" of AI, then computing power is the "circulatory system" that supports the operation of the brains. NVIDIA knows well that controlling computing power means controlling the lifeline of AI. So, it has started to make crazy layouts in the field of computing power infrastructure.

The story of Nscale is quite dramatic. This company was spun off from an Australian cryptocurrency mining company in 2023 and transformed into an AI data center. In September 2025, it completed an $1.1 billion financing; then in October, NVIDIA participated in its $433 million SAFE financing.

Why does NVIDIA value Nscale so much? Because it is building data centers in the UK and Norway for OpenAI's "Stargate Project". Stargate is OpenAI's super - computing power plan, which requires deploying a huge number of GPU clusters globally. Nscale is responsible for building the "roads", OpenAI is responsible for "driving the cars", and NVIDIA provides support for both sides.

Crusoe has a different business model. This company's business model is straightforward: it builds data centers and then rents them out to giants such as Oracle, Microsoft, and OpenAI. In November 2024, Crusoe completed a $686 million financing round, led by the Founders Fund, and NVIDIA was one of the co - investors.

There is an interesting phenomenon here: NVIDIA invests in both "users" like OpenAI and "landlords" like Crusoe. Users need computing power, landlords provide computing power, and the core of computing power - GPUs - comes from NVIDIA. This is a perfect closed - loop.

CoreWeave is the most successful case in this story. When NVIDIA invested in April 2023, CoreWeave was still a startup, and the financing amount was $221 million. But just two years later, CoreWeave has successfully gone public and become a benchmark enterprise in the GPU cloud service field. NVIDIA is still an important shareholder to this day.

There are also Lambda and Together AI. Lambda completed a $480 million Series D financing in February 2025, with a valuation of $2.5 billion. Its main business is to rent NVIDIA GPU servers to AI companies for model training. Together AI completed a $305 million Series B financing in February 2025, with a valuation of $3.3 billion, and it provides cloud infrastructure for AI model building.

These companies have a common feature: their businesses are highly tied to NVIDIA. The more successful they are, the more NVIDIA GPUs they need to purchase; the more customers they have, the deeper the entire AI ecosystem's dependence on NVIDIA.

AI also has a fatal problem: power consumption. Training a large model may consume as much electricity as a small city. If the energy problem is not solved, the development of AI will eventually encounter a bottleneck.

NVIDIA clearly realizes this. In August 2025, it participated in Commonwealth Fusion's $863 million financing round. This company is researching nuclear fusion energy - a technology that can theoretically provide almost infinite clean energy. Co - investors include Google and Bill Gates' Breakthrough Energy Ventures, which shows the importance of this project.

If nuclear fusion is successfully commercialized, the energy problem of AI will be completely solved. And NVIDIA has already secured its position in advance.

Firmus Technologies has taken a more practical approach. This company originally focused on bitcoin mining cooling technology and then made a magnificent transformation. It is building an "AI factory" in Tasmania, Australia, focusing on high - efficiency AI infrastructure. In September 2025, it completed a A$330 million (about $215 million) financing round, and its valuation reached A$1.85 billion (about $1.2 billion). NVIDIA is one of the co - investors.

From bitcoin to AI, from mining to computing power, Firmus' transformation has exactly caught up with the trend of the times. And NVIDIA can always find its place in these transformation stories.

In addition to energy, NVIDIA is also betting on more fundamental technological breakthroughs. Ayar Labs is a typical example. This company develops optical interconnection technology, which can significantly improve the data transmission efficiency between AI chips while reducing energy consumption. NVIDIA has invested in this company three times, with the latest being a $155 million financing in December 2025.

Investing in the same company three times shows that NVIDIA is highly optimistic about this technology direction. Optical interconnection may be the key technology for the next - generation AI chips, and NVIDIA wants to ensure that it does not fall behind in this field.

From data centers to energy technologies, from nuclear fusion to optical chips, NVIDIA is building a complete computing power ecosystem. It is not just selling GPUs but building "highways" for the entire AI era. And when everyone is running on these "highways", the one collecting the tolls will always be the winner.

03

No matter how advanced the technology is, if it cannot be commercialized, it will ultimately be a castle in the air. NVIDIA understands this well, so in its investment portfolio, there are many companies focusing on AI application scenarios. From autonomous driving to robots, from healthcare to content creation, NVIDIA wants to make AI ubiquitous.

Autonomous driving is one of the most attractive and capital - intensive application scenarios of AI. NVIDIA's layout in this field reflects its long - term vision.

Wayve is a UK - based company that develops self - learning autonomous driving systems. In May 2024, NVIDIA participated in its $1.05 billion financing round; in September 2025, NVIDIA plans to invest an additional $500 million. Wayve is currently conducting road tests in the UK and the San Francisco Bay Area. Its technology route is to enable vehicles to master driving skills through "self - learning" rather than relying on high - precision maps.

Waabi focuses on autonomous trucks. In June 2024, it completed a $200 million Series B financing round, led by Uber and Khosla Ventures, and NVIDIA was one of the co - investors. Interestingly, the investors in this financing round also included Volvo and Porsche, which shows that traditional automakers are also betting on this direction.