Data InsightsNVIDIA’s revenue from data centers and AI has grown 1,300-fold in the last 12 years

NVIDIA’s revenue from data centers and AI has grown 1,300-fold in the last 12 years

Stacked bar chart of NVIDIA's quarterly revenue by market segment from February to April 2014 through February to April 2026 where data-center and AI revenue has grown about 1,300-fold in 12 years to represent over 90% of the total by 2026. Data source: NVIDIA Corporation (2026). License: CC BY.

Spending on the hardware that trains and runs artificial intelligence has grown rapidly over the past years. One of the clearest indicators of how rapidly spending has increased is the revenue of American chipmaker NVIDIA.

The company accounts for around 85% of the global market for AI chips. These are graphics processing units (GPUs), originally built for video games but well-suited to the parallel computation AI training requires.

The chart shows NVIDIA's quarterly revenue in US dollars, split by end market — sales to data-center customers (cloud providers and AI companies) in green, and sales for gaming, consumer devices, and cars in red.

In early 2014, data centers and AI accounted for just 5% of its revenue; gaming was the biggest single segment. Twelve years later, the ratio has flipped: data centers and AI now make up over 90% of revenue. The revenue in this segment has grown 1,300-fold over the period, from $57 million to more than $75 billion per quarter.

The data centers and AI segment was already growing fast between 2014 and 2022, with revenue doubling every 16 months on average. ChatGPT's release in late 2022, alongside the broader push to deploy AI at scale, has accelerated that pace: since then, revenue has doubled every 11 months.

Read our article on how increasing compute, data, and model size has made artificial intelligence more capable

Our latest Data Insights

See all Data Insights