Nvidia pushes broad ecosystem thinking

Nvidia is making a case that the artificial intelligence market will be built by many companies, not one champion, and it is investing accordingly. (tekedia.com) Jensen Huang said on the Dwarkesh podcast, in comments published April 16, that Nvidia tries to invest across foundation-model companies because it does not want to “pick winners.” He tied that view to Nvidia’s own early history, when he said about 60 graphics companies were competing in the 1990s. (tekedia.com) Nvidia’s balance sheet gives that strategy unusual weight. The company reported fiscal 2026 revenue of $215.9 billion on November 19, 2025, including $68.1 billion in fourth-quarter revenue and $62.3 billion from data center sales in that quarter alone. (investor.nvidia.com) The company is not just selling chips into that market. Nvidia said in a December 11, 2023 post that it had made more than two dozen ecosystem investments that year through corporate development, NVentures, and its startup programs. (blogs.nvidia.com) That helps explain why Nvidia keeps showing up in places that look far from the chatbot race. NVentures says it backs companies in robotics, biotech, enterprise software, energy, autonomous systems, and industrial tools, alongside model and infrastructure startups. (nventures.ai) Nvidia has also built programs meant to widen the funnel. Its Venture Capital Alliance says it connects investors to more than 30,000 AI startups through Nvidia’s network, while the Inception program offers cloud credits, developer tools, training, and go-to-market support. (nvidia.com) The same logic shows up in infrastructure deals. Nvidia said on March 11, 2026 that it would invest $2 billion in Nebius and support deployments of more than 5 gigawatts of Nvidia systems by the end of 2030. (investor.nvidia.com) CNBC reported on January 26, 2026 that Huang described Nvidia’s separate $2 billion CoreWeave investment as only a “small percentage” of the capital still needed to build out that scale of computing capacity. CoreWeave disclosed a $6.3 billion order from Nvidia in September, according to the report. (cnbc.com) For Bay Area real estate and office markets, that points to a wider set of possible tenants than model labs alone: cloud operators, robotics firms, simulation software teams, biotech platforms, and industrial AI companies that use Nvidia’s stack. Nvidia’s own investment and partner programs are already organized around that broader mix. (nventures.ai) (nvidia.com) Huang’s argument is simple: if the field stays fragmented, Nvidia can still sit underneath much of it. The company’s recent investments, startup programs, and cloud partnerships show it is spending as if that outcome is the base case. (tekedia.com) (investor.nvidia.com)