OpenAI has formalized a contract with Amazon Web Services valued at 38.000 million (around 33.000 billion euros) to ensure cloud computing power during seven yearsThe announcement was made via a joint statement and was accompanied by Amazon shares rise in the opening moments of the session.
The agreement opens OpenAI's access to hundreds of thousands of Nvidia GPUs, including GB200 and GB300organized in low-latency clusters using Amazon EC2 UltraServers. The roadmap anticipates immediate capacity deployment and having full operational deployment before the end of 2026, with the option to extend the collaboration from 2027.
Keys to the agreement and roadmap
In the first phase, the company will use the existing AWS data centersNext, Amazon will lift specific infrastructure for OpenAI with an architectural design intended to maximize efficiency and performance in demanding AI workloads.
OpenAI emphasizes that scaling the AI ​​frontier requires massive and reliable computingand that this alliance strengthens a broad computing ecosystem to bring advanced capabilities to more users. For their part, AWS emphasizes that its first-class infrastructure It is designed to handle extremely complex workloads with security, performance, and scalability.
Competitive context and governance changes
This move comes after the renegotiation of the agreement with Microsoftwhich until 2023 had been OpenAI's exclusive cloud provider. With the new framework, the company can contract services on the open market, thus adopting a multi-vendor strategy to avoid bottlenecks.
In addition to AWS, OpenAI has closed collaborations with Oracle infrastructures , Google Cloud, CoreWeave and key manufacturers such as Nvidia o AMDIndustry analysts point out that these increasingly interconnected relationships They increase investment intensity and fuel the debate about a possible bubble risk in the AI ​​ecosystem.
In parallel, OpenAI has implemented changes in its structure to operate with profit motive And, as Reuters reported, it paves the way for a potential initial public offering with very high valuation expectations. All of this fits into an expansion phase in which the company is seeking large-scale computing power to train and deploy new models.
Relevance for Europe and Spain
For the European market, the agreement could translate into greater availability of resources in AWS regions within the EU and Spain, facilitating regulatory compliance (such as GDPR) and providing lower latency for companies that integrate generative AI solutions.
The expansion of computing capacity is often accompanied by debates about energy, sustainability and permits around data centers. In Europe, these regulatory and infrastructure considerations are keys to deployment of large-scale AI projects.
For Spanish companies in regulated sectors—finance, health, public administration or industry—having contained latency and a robust platform can accelerate AI pilots and production deployments, provided the following conditions are met security and compliance requirements.
Market and technology
The announcement had an immediate impact on the stock markets: Amazon advanced by about 5% In the early stages of trading, stocks and values ​​linked to AI hardware also reacted positively. These are short-term movements that the market typically associates with deals of this size due to their potential operational impact.
On a technical level, the clusters with Next-generation GPU will allow OpenAI to tackle both training and large-scale inference, with low latency and high performance in interconnected networks. The combination of EC2 UltraServers and the AWS network fabric is designed to maximize efficiency in intensive workloads and improve the security of data and models.
The alliance between OpenAI and AWS articulates a capacity jump which seeks to consolidate the supply of computing power for the coming years, setting ambitious deployment timelines —full capacity before 2026— and strengthen cloud competition, with direct implications for the business fabric of Spain and Europe.
