Nvidia, OpenAI and nine others pledge $2.4 billion in AI compute for US science

Eleven technology companies, led by Nvidia, have pledged about $2.4 billion in computing power, cloud access, AI models and related support for the Genesis Mission, the US government's programme to use artificial intelligence in scientific research. The commitments were announced at a White House science summit on Thursday, October 8, 2026, as part of a wider science package the administration valued at more than $6 billion. Nvidia pledged $1 billion, AMD $500 million and OpenAI $200 million, with Anthropic and Google committing $150 million each. The US Department of Energy separately announced $159 million for 12 new Phase II Genesis Mission projects.

Modern scientific research, from drug discovery to materials and climate modelling, increasingly depends on huge amounts of computing power and access to the most capable AI models. Both are expensive and concentrated in a handful of companies. The Genesis Mission is Washington's attempt to channel that private capacity into government science, and this round of pledges is the clearest sign yet that the biggest AI and chip firms are willing to take part.

It is worth being precise about what was promised. Most of the $2.4 billion is in kind: credits on cloud platforms, time on chips, access to AI models and discounts. That is valuable to researchers who would otherwise wait in line for computing time, but it is not the same as new public funding, and the real value will depend on how easily scientists can use what has been offered.

For India, the announcement is a reminder of how closely AI capability is now tied to national science strategy. India's own IndiaAI Mission has focused on building shared computing capacity and supporting home-grown AI models for Indian languages and needs. The US approach shows another model: getting private companies to commit capacity directly to public research. Indian policymakers, universities and research labs will watch whether such public-private compute deals speed up scientific results, and whether a similar approach could work with companies operating in India.

The list of pledgers also shows how the AI industry is lining up. Chip makers Nvidia, AMD and Micron sit alongside model builders OpenAI, Anthropic and Google, cloud provider AWS and newer infrastructure firms such as Crusoe and Armada. All of them benefit from showing that their technology serves public goals as governments debate how to regulate AI.

Short-term: US federal science agencies and national laboratories gain access to more computing time, cloud credits and AI models for research projects, including the 12 new Phase II projects backed by the Department of Energy.

Long-term: If the arrangement produces faster scientific results, it could become a template for how governments and AI companies share capacity for public research, a model other countries, including India, may study.

Who is affected: US government scientists and research labs, the 11 pledging companies, and researchers and policymakers in other countries, including India, who are planning national AI compute strategies.

Questions and Answers