GitHub added 4 million CPU cores to absorb AI-generated code

GitHub says agentic AI drove a sixfold year-on-year increase in new code. To cope, it added 4 million CPU cores and 160 petabytes of storage to its data centers.

Why it matters: AI coding agents are now a measurable infrastructure load, not just a productivity tool. Code hosting, CI and review pipelines must scale with machine-generated volume, which shifts AI demand beyond GPUs toward CPUs and storage, a trend also visible in the push to benchmark CPUs for agentic workloads.