Nvidia, the world’s most valuable semiconductor company, has dramatically accelerated its push beyond chips and into venture investing, closing at least 67 AI-linked startup deals in 2025 as it cements its position at the center of the global artificial intelligence boom.
The scale of Nvidia’s investment activity this year already surpasses the 54 deals it completed in all of 2024, according to PitchBook data, underscoring how the company is increasingly using capital, alongside its dominant GPUs, to shape the direction of the AI industry.
The tally excludes investments made through NVentures, Nvidia’s formal corporate venture arm, which has also sharply increased its pace, participating in about 30 deals in 2025 compared with just one in 2022.
The surge comes amid Nvidia’s extraordinary financial ascent. Since the launch of ChatGPT and the rapid spread of generative AI services, Nvidia’s revenue, profitability, and cash reserves have soared, propelling its market capitalisation to about $4.6 trillion.
The company has redeployed that windfall into a sprawling network of startups spanning foundation models, AI infrastructure, robotics, autonomous driving, data centres, healthcare, and energy.
Nvidia has said its corporate investing strategy is aimed at expanding the AI ecosystem by backing companies it views as game changers and market makers.
In practice, the approach has placed the chipmaker at the intersection of many of the industry’s biggest funding rounds, often in deals that also lock startups into long-term demand for Nvidia hardware.
At the top end of Nvidia’s 2025 dealmaking are multi-billion-dollar rounds involving the most influential names in artificial intelligence.
In October 2024, Nvidia made its first direct investment in OpenAI, reportedly committing $100 million as part of a $6.6 billion round that valued the ChatGPT maker at $157 billion. While Nvidia did not participate in OpenAI’s subsequent $40 billion funding round, the company later announced a broader strategic partnership under which it could invest up to $100 billion over time to deploy massive AI infrastructure, though it cautioned in filings that there is no assurance such investments will ultimately be completed.
Read also: Sora secures $2.5m more in seed funding to fight malaria in Africa
A similar pattern of circular capital flows appears across Nvidia’s largest bets. In November 2025, the company committed up to $10 billion to Anthropic in a strategic round that also included a $5 billion investment from Microsoft. As part of the arrangement, Anthropic committed to spending roughly $30 billion on Microsoft Azure compute while also purchasing Nvidia’s next-generation Grace Blackwell and Vera Rubin systems.
Nvidia has also backed many of the fastest-growing AI startups of the past year. It participated in Cursor’s $2.3 billion Series D round, which valued the AI coding assistant at $29.3 billion, and joined the $6 billion financing of Elon Musk’s xAI, with plans to invest up to $2 billion more to help the company acquire Nvidia hardware.
In Europe, Nvidia invested for the third time in France-based Mistral AI, backing its €1.7 billion Series C at a post-money valuation of about $13.5 billion.
Beyond model developers, Nvidia has spread its capital across the infrastructure that underpins the AI boom.
It invested in Crusoe’s $1.4 billion Series E to build data-centre campuses in Texas and Wyoming for Oracle-leased facilities powering OpenAI workloads, and backed Nscale, which is constructing data centres in the U.K. and Norway for the same Stargate project.
Nvidia has also invested in GPU cloud providers such as CoreWeave, Together AI, and Lambda, many of which rent out servers built around Nvidia chips. The company’s reach extends into robotics and autonomous systems.
Nvidia participated in Figure AI’s $1 billion-plus Series C round, valuing the humanoid robotics startup at $39 billion, and invested in Wayve, Nuro, Waabi, and Bright Machines, all of which are building AI-driven systems for transportation and manufacturing.
Healthcare, energy, and scientific computing have also featured prominently. Nvidia backed Hippocratic AI, which is developing healthcare-focused large language models, Commonwealth Fusion Systems in nuclear fusion, and Sandbox AQ, which is building large quantitative models for complex numerical analysis. It has also invested in data-labelling company Scale AI, optical interconnect startup Ayar Labs, and enterprise AI platforms such as Cohere, Uniphore, and Kore.ai.
Taken together, the 67 deals illustrate how Nvidia’s influence now extends far beyond supplying GPUs. By investing directly in startups that both depend on and help expand demand for AI compute, the company has positioned itself at the centre of what analysts increasingly describe as a circular AI economy, where capital, hardware sales, and infrastructure spending reinforce one another.
That structure has drawn growing scrutiny. Separate analysis of AI infrastructure flows suggests that by 2027, OpenAI’s projected annual infrastructure commitments to companies such as Oracle, Nvidia, AMD, and Broadcom could exceed its projected revenue by more than three times, highlighting how continued external capital may be required to sustain current growth trajectories.
For Nvidia, however, the strategy has so far proven effective. Its venture investments deepen customer relationships, accelerate adoption of its latest chips, and give the company early visibility into emerging AI applications. As long as demand for compute continues to surge, Nvidia’s blend of silicon and capital appears set to remain one of the most powerful forces shaping the AI industry.
Whether that investment-driven expansion ultimately proves sustainable for the broader ecosystem remains an open question, but in 2025, Nvidia has made clear it intends to be everywhere AI is being built.
