Not just selling chips! NVIDIA is transforming into the strongest AI venture capital, having invested in nearly 60 companies this year

NVIDIA (NVDA.US), although not a traditional venture capital institution, has quietly become one of the most influential participants in the venture capital field.

According to the Zhitong Finance APP, this company, with the highest market value globally, announced in September plans to invest $100 billion in OpenAI—this deal will become the largest startup investment in history. However, this is just one part of NVIDIA's expanding investment layout in private artificial intelligence (AI) companies worldwide.

Data from PitchBook shows that as of mid-October, this chip manufacturer has invested in 59 AI startups this year, betting on at least 10 companies in just the past two months. This number has surpassed its record of 55 investments for the entire year of 2024 and represents a significant increase from just 12 investments in 2022.

NVIDIA CEO Jensen Huang stated that the company's strategy is to support a broader AI ecosystem. Currently, NVIDIA is the biggest beneficiary of the surge in AI infrastructure spending. Analysts point out that to maintain this growth trend, it is essential to promote the application of AI technology across a wider range of economic sectors, and the new application scenarios developed by startups will be key. By supporting emerging AI companies, NVIDIA is following the conventional path of corporate venture capital departments—but at a faster pace and with higher stakes.

Bernstein analyst Stacy Rasgon stated, "Clearly, as the AI ecosystem expands, NVIDIA wants all constructions to be based on its technological architecture. Once an ecological barrier is formed, it will be difficult for competitors to shake its position."

While continuing to increase its investment in the AI field, Jensen Huang emphasized that he does not believe there is a bubble in the AI industry. "Like many other companies, NVIDIA is simply betting that the AI industry will eventually become profitable," Rasgon commented.

In addition to its significant investment in OpenAI, NVIDIA has also supported various AI companies of different sizes. These include Reflection AI, which develops autonomous programming agents, Reka AI, which researches video analysis models, and Perplexity AI, an AI search engine company aimed at challenging Google's (GOOGL.US) position.

For these startups, NVIDIA's investment not only brings capital support but also provides access to its management and scarce computing resources. Perplexity CEO Aravind Srinivas revealed in an interview last year that he had requested technical resources from NVIDIA executives to optimize product demonstrations, and Jensen Huang instructed the team to provide support, resolving the request within half an hour.

"I have made similar requests to CEOs of other companies, and those requests usually get passed to department heads or vice presidents first, and then they would arrange another meeting to further understand the specific use case. But with Jensen Huang, things can be settled quickly," Srinivas said Some startups are using funds to purchase NVIDIA's high-performance chips, which are essential for training cutting-edge AI models and cost tens of thousands of dollars each. Although the computing power requirements vary among companies, most plan to stockpile thousands of processors. NVIDIA does not require its invested companies to purchase its products, but these chips often become one of the main expenditure items for startups.

Taking Reka as an example, it completed a $110 million financing this year with support from NVIDIA. CEO Dani Yogatama stated that part of the funds will be used to enhance computing power. The company has also joined NVIDIA's "Inception" program, which empowers AI startups through a technology platform and commits to deep collaboration with the chip manufacturer. Yogatama pointed out that access to NVIDIA's engineers and distribution network is significant for Reka's customer expansion and product improvement.

Yogatama said, "Companies like NVIDIA have a wide influence globally and can help us promote technology. For small companies like ours, even if we develop the most advanced technology, if no one uses it, it cannot realize its value."

Undoubtedly, NVIDIA occupies a unique ecological niche: its dominance in the high-end chip market has propelled the company's market value to $4.9 trillion. The surge in sales has brought abundant free cash flow. Morningstar analyst Brian Colello noted, "NVIDIA's cash will be so abundant that it won't know how to dispose of it."

Colello also stated, "Given that many transactions may not pass the scrutiny of the U.S. Federal Trade Commission (FTC), making smaller investments at this stage is more reasonable for the company."

Investing in startups is expected to help NVIDIA diversify its customer base. Currently, large tech companies still account for a disproportionately high share of the company's clientele. The top five customers contribute nearly 50% of revenue: including Microsoft (MSFT.US), Meta (META.US), Amazon (AMZN.US), and Google (GOOGL.US) — these four major cloud service providers heavily utilize NVIDIA chips in their data centers; and Supermicro (SMCI.US) — whose servers were equipped with NVIDIA chips before entering data centers. Although not yet profitable, OpenAI, which has committed to spending trillions of dollars on AI infrastructure, is also seen as one of its major customers.

At the same time, many large enterprises are working to reduce their dependence on NVIDIA chips. On October 6, OpenAI reached a partnership with NVIDIA competitor AMD (AMD.US) to deploy AMD chips worth hundreds of billions of dollars. The following week, OpenAI also agreed to procure custom chips from Broadcom (AVGO.US) to provide computing power for ChatGPT. Google is also collaborating with Broadcom to develop self-researched chips to help reduce costs for its Gemini model Despite some analysts' concerns about the cyclical dependence within the AI ecosystem, NVIDIA's continued investment in AI companies reflects an optimistic judgment about the industry's prospects in the long run. Recently, the wave of transactions and collaborations involving NVIDIA and a few core AI players has intensified market worries—whether the increasingly complex and intertwined business networks are artificially supporting the trillion-dollar scale of AI prosperity.

However, Jensen Huang firmly believes that the industry is still on the brink of an explosion. "We have made several outstanding investments," the CEO said in an interview this month, "If there is any regret, it would probably be that we should have invested more back then."