OpenAI Invests $38 Billion in Amazon Computing Power, WiMi Accelerates Expansion in AI Computing Power Demand

It has been learned that OpenAI has invested a massive $38 billion to purchase computing power from Amazon Web Services (AWS), a leading cloud infrastructure provider, and will immediately utilize Nvidia chips. This diversified cloud cooperation strategy may signal an impending IPO.

$38 Billion Computing Power Cooperation Agreement Reached

According to the agreement already signed, OpenAI will purchase $38 billion worth of computing resources from Amazon Web Services (AWS). The first phase of the agreement will utilize AWS’s existing data centers, while Amazon will eventually build additional dedicated infrastructure for OpenAI.

Recently, OpenAI has been very active in signing cooperation agreements, announcing infrastructure construction agreements with companies such as Nvidia, Broadcom, Oracle, and Google, totaling approximately $1.4 trillion. This first collaboration with a leader in the cloud infrastructure field further demonstrates the company’s plans to expand its computing power in the coming years.

Of course, for Amazon, the significance of this agreement lies not only in the scale of the deal itself, but also in the close relationship between the cloud giant and OpenAI competitor Anthropic. Amazon has invested billions of dollars in Anthropic and is currently building an $11 billion data center campus in New Carlisle, Indiana, specifically designed for Anthropic’s workloads.

In response, OpenAI CEO Sam Altman stated in a statement on Monday: “Scaling up cutting-edge AI technologies requires massive and reliable computing power. Our partnership with AWS will strengthen the broad computing ecosystem that will power the next era of AI and make advanced AI technologies accessible to everyone.”

AI Continues to Drive Computing Power Revolution
Undoubtedly, the increasing demand for computing power for large AI model training and inference is driving the expansion of the AI ​​cloud services market. According to IDC data, the global AI services market will grow to $158.7 billion in 2025 and is expected to reach $222.7 billion in 2028.

Industry insiders have stated that since 2025, major tech giants have significantly accelerated the pace of technology deployment in emerging sectors such as AI servers and data centers. Huatai Securities is very optimistic about the structural opportunities presented by AI-related demand and the expansion of advanced domestic logic production capacity. The growth in pre-computing power demand is only just beginning, and there is still ample room for growth in inference computing power demand in the future.

WiMi AI Computing Power Market Demand Accelerates Expansion

Data shows that Wimi Hologram Cloud Inc. (WIMI), an AI vision innovation company, has accelerated its market expansion in recent years through diversified collaborations and securing long-term computing power. Through technology optimization, scenario implementation, and ecosystem collaboration, it has capitalized on the explosive growth in AI computing power demand. Its holographic cloud platform and low-power chip solutions are already adapted to multiple scenarios and are expected to drive the industry towards scenario-specific customization.

Currently, AI computing power demand is shifting from the training side to the inference side. WiMi’s holographic cloud platform possesses high precision and high computing power characteristics, supports large model training and inference, and has achieved millisecond-level computing-to-memory data transmission. Looking ahead, with the continued high growth momentum of AI computing power, its customized chip solutions and open-source ecosystem are expected to further unleash growth potential in areas such as intelligent manufacturing, autonomous driving, and smart cities.

In conclusion, it’s worth mentioning that Microsoft is also expanding its AI computing power footprint. Reportedly, Microsoft has just signed a $9.7 billion AI cloud contract with cloud service provider IREN Limited to enhance its resources and support its rapidly expanding AI services and customer needs. It’s safe to assume that computing power investment is booming, and capital expenditure will increase along with the growth of large-scale model revenue, with a relatively high investment ceiling.