IDC: Arm Servers Surpass x86, AI Infrastructure Spending to Reach $497 Billion by 2026
N.R. Finch
IDC data shows Arm-based GPU servers surpassed x86 in accelerated-computing market share for the first time in Q1 2026, while full-year AI infrastructure spending is projected at $497 billion — the data-center compute stack is being rebuilt.
How big is the spending — and where is it going?
Global AI infrastructure spending hit $89.7 billion in Q1 2026.
IDC projects the full-year figure at $497 billion. This means → AI hardware investment is approaching the scale of total global semiconductor revenue — no longer a line item inside IT budgets, but a standalone spending category.
Why did Arm overtake x86?
Among accelerated-computing platforms — rack-scale servers packed with GPUs — Arm-based servers took majority market share for the first time, displacing x86 as the dominant architecture.
In plain terms = for decades, nearly every data-center server ran on Intel or AMD x86 chips; the new wave of AI-accelerator servers is mostly built on Arm.
This reflects AI training and inference's extreme demand for performance per watt — Arm's inherently lower power draw gives it a structural edge in GPU-dense racks.
Who faces pressure — and who benefits?
Intel bears the most direct hit: x86 servers lost their dominant position in the fastest-growing segment of data-center compute.
AMD sits in the same x86 camp, but its GPU business (MI series) partially offsets the CPU-side share loss.
This means → the data-center compute landscape is undergoing a structural shift — not a one-quarter blip, but a generational change in the underlying instruction-set architecture. For investors, the old equation "server = x86" no longer holds.
Content is for reference only, not financial advice.