NextFin News - OpenAI CEO Sam Altman publicly expressed his gratitude to Nvidia CEO Jensen Huang this week, marking a symbolic high point in a $100 billion infrastructure alliance that is redrawing the map of Silicon Valley power. The public display of camaraderie follows a series of high-stakes negotiations that have effectively ended Microsoft’s era as the exclusive gatekeeper of OpenAI’s computing needs. By securing a direct pipeline to Nvidia’s upcoming Vera Rubin platform, Altman has ensured that the next generation of frontier models will not be throttled by the capacity constraints that plagued the industry throughout 2024 and 2025.
The scale of the collaboration is staggering. Under the terms of the agreement, OpenAI is set to deploy at least 10 gigawatts of Nvidia-powered systems, a massive buildout that Huang has described as the largest AI infrastructure project in history. To put that in perspective, 10 gigawatts is roughly equivalent to the power output of seven or eight large nuclear power plants, all dedicated to the singular task of training and running artificial intelligence. Nvidia is not merely a vendor in this arrangement; it is a primary financier, committing to invest up to $100 billion in OpenAI in tranches as each gigawatt of capacity comes online.
This deepening bond between the world’s most valuable chipmaker and its most prominent AI lab has created palpable friction with legacy partners. According to reports from CNBC, Microsoft was only informed of the definitive deal a day before it was signed, a move that underscores Altman’s aggressive pursuit of "compute sovereignty." While Microsoft remains a major shareholder, the exclusivity that once defined the relationship has evaporated. Huang himself reportedly took the lead in informing Microsoft CEO Satya Nadella of the shift, signaling that Nvidia is now comfortable playing the role of both supplier and kingmaker in the cloud ecosystem.
The timing of Altman’s public thanks coincides with the integration of OpenAI’s projects into the "Stargate" umbrella, a multi-billion dollar infrastructure initiative backed by U.S. President Trump’s administration along with Oracle and SoftBank. By aligning with Nvidia, Altman is insulating OpenAI from the risk of being tied to a single cloud provider’s hardware roadmap. The first gigawatt of the new Vera Rubin systems is expected to begin generating tokens in the second half of 2026, promising a billion-fold increase in computational power compared to the original DGX server Huang hand-delivered to OpenAI a decade ago.
For Nvidia, the $100 billion commitment to OpenAI is the crown jewel of a broader strategy to vertically integrate the AI economy. Beyond OpenAI, Huang has recently directed a $5 billion investment into Intel for data center chip co-development and nearly $700 million into the U.K. data center startup Nscale. These moves suggest that Nvidia is no longer content just selling shovels; it is now building the mines and owning a stake in the gold. The risk, of course, is that by becoming so deeply embedded with OpenAI, Nvidia may alienate other major customers like Google or Amazon who are racing to develop their own internal silicon.
The geopolitical dimension cannot be ignored. With U.S. President Trump’s administration pushing for domestic AI supremacy through the Stargate project, the Altman-Huang alliance serves as the industrial backbone of national policy. The sheer energy requirements of 10 gigawatts will require unprecedented coordination with utility providers and federal regulators. As the first clusters of the Vera Rubin chips prepare for installation, the gratitude Altman expressed this week reflects a pragmatic reality: in the race for artificial general intelligence, the winner will be the one who controls the most electricity and the most silicon. For now, those keys are held firmly by Jensen Huang.
Explore more exclusive insights at nextfin.ai.
