According to the International Data Corporation (IDC) Worldwide Quarterly Server Tracker, the server market reached a record $95,2-billion dollars in market value during the first quarter of the year, showing the highest quarterly growth rate ever with 134,1% increase compared to Q1 2024.
This outstanding performance pushed the market projection up for the year to a record $366-billion dollars for 2025 which represents a 44,6% growth compared to 2024.
The x86 server market value is projected to increase 39,9% in 2025 to $283,9-billion while Non-x86 servers will increase 63,7% year over year to $82-billion.
Servers with an embedded GPU are projected to grow 46,7% year-over-year representing almost 50% of the total market value in the year.
The fast pace at which hyperscalers and cloud service providers continue adopting servers with embedded GPUs has fueled the server market growth which will be tripling in size in only three years.
Thanks mainly to the launch of large rackscale configurations ARM based servers are rapidly gaining traction in the market by growing at 70% and representing 21,1% of the total shipments of the year.
“The Stargate project re-announcement promised to invest up to $500-billion in AI infrastructure to help create artificial general intelligence (AGI). Shortly thereafter, the release of DeepSeek’s R1 reasoning model caused concerns about the necessity of investing in so much infrastructure,” says Kuba Stolarski , research vice-president: Worldwide Infrastructure Research at IDC.
“R1 needed more infrastructure than was reported, and the evolution from simple chatbots to reasoning models to agentic AI will require several orders of magnitude more processing capacity, especially for inferencing. Improvements in the efficiency of model creation were expected and in fact a goal in the industry. Efficient models will use fewer resources, and therefore may scale better in multi-user environments, enabling high level reasoning and possibly eventually leading to AGI.”
The US is the fastest growing region in the server market with an increase of 59,7% compared to 2024, representing almost 62% of the total revenue in 2025. China is also growing at a faster rate than other regions with 39.5% year-over-year growth in 2025 and accounting for more than 21% of the quarterly revenue worldwide. Japan, and APeJC had double-digit growth projected this year with 33,9% and 10,8% respectively. EMEA and Latin America showed single digit growth at 7% and 0,7%, while Canada is showing negative -9,6% growth due to an unusual very large deal in 2024.
Definitions
IDC defines a server system as a multiuser computing device that accesses and delivers services via a network. The server and the applications that run on it are typically shared by multiple users.
Unlike a client device, a server does not usually have a user interface that is intended for human-machine interaction. A typical server system entails one or more processors, a motherboard, memory, internal disk or flash storage, a bundled operating system (OS), power supply units, and network interfaces.
Accelerator Type Definitions
Non-accelerated: These servers do not have an embedded accelerator. Servers where the accelerator is added by end users that have purchased the acceleration technology directly from an acceleration vendor and have installed it themselves are considered non-accelerated in IDC’s Server Tracker. Servers with an integrated graphics processor (IGP) (i.e., fused to the motherboard) are also considered non-accelerated in IDC’s Server Tracker.
GPU (graphics processing units): A GPU is a processor specialized for rendering images, animations, or video to a computer display. In the context of server-based accelerated computing, a GPU is typically a programmable discrete (removable) graphics card but may also be an integrated graphics processor (i.e., fused to the motherboard). In addition to its graphical applications, a GPU may also be used to perform operations traditionally handled by a central processing unit, a concept known as a general-purpose graphics processing unit (GPGPU).
Other accelerated: This category combines servers with a discrete embedded accelerator FPGA or ASIC. A field-programmable gate array is an integrated circuit designed to be configured by a customer after manufacture. FPGAs are made up of an array of programmable logic blocks, interconnects, and input/output (I/O) blocks. An application-specific integrated circuit is a purpose-built integrated circuit that cannot be reconfigured after manufacture.