Tech Giants are Going to Spend More on AI Hyperscalers Because of Increasing Demand

Hyperscalers are large scale data center operators which provide services in cloud computing and storage and the data by J.P Morgan Asset Management shows that spending for AI hyperscalers is going to increase in the upcoming years. Companies like Alphabet/Google, Microsoft, Meta, Amazon and Oracle are spending a lot on AI hyperscalers because they are important for training AI models. These companies have spent estimated $197 billion on AI infrastructure in 2024 and the total capital spending for hyperscalers is set to rise further.

The highest spending on hyperscalers was by Microsoft, a total of $75 billion in 2024. Amazon Web Services (AWS) is also increasing its investments in AI infrastructure because of global demand for advanced applications. Meta spent about $40 billion on AI hyperscalers in 2024, with an increase of 350,000 GPUs to train their LLMs. Google Cloud is also investing a lot in training their AIs which is making them billions of revenue. The need for data centers is rising with increase in demands.

It is also expected that tech giants will spend $234 billion collectively on AI hyperscalers in 2025, while about $248 billion will be spent in 2026. It is also to be noted that US data center property inventory has increased by 43% in 2023 and 2024 because there is a high demand from big tech firms.


YearAI Hyperscaler Capital Spending
2026P$249 billion
2025P$234 billion
2024P$197 billion
2023$126 billion
2022$123 billion
2021$92 billion
2020$73 billion
2019$67 billion
2018$65 billion
2017$40 billion
2016$31 billion
2015$25 billion
2014$23 billion
2013$17 billion
2012$10 billion
2011$8 billion
2010$7 billion
2009$4 billion
2008$7 billion
2007$5 billion
2006$4 billion
2005$2 billion
2004$2 billion

Read next: New Report Highlights AI Challenges, Expanding Applications, and Emerging Competitors Shaping 2025's AI Landscape
Previous Post Next Post