جهش شش‌برابری کدهای هوش مصنوعی در گیت‌هاب و نیاز به چهار میلیون هسته پردازشی جدید

GitHub says massive influx of AI-generated code required extra 4 million CPU cores and 160 petabytes of storage to manage — a sixfold increase to 7.4 billion monthly commits in one year

Everyone knew the rise was sharp, but the figures are staggering.

Is most programming code now made by AI? That's not hard to believe given the number of testimonies, but we now have exact figures behind that statement. GitHub, the world's largest code repository, says that over the course of this year, it added 4 million CPU cores, 160PB of storage, and 50 Tb/s of networking capacity to keep up with the influx of AI-generated code.

Raw repository activity saw a massive year-over-year increase, as per GitHub's technical report. Code commits (adding/removing code) grew sixfold, while merged pull requests (finalizing a feature or bugfix) jumped 4x.

The multipliers sound impressive enough, but the exponential growth pattern is even more so: half a year ago in April, there were already 1.4 billion commits per month — a figure that rose to 2.9 billion in August, and now sits at 7.4 billion for September. GitHub saw 17.9 billion code pushes in the past 12 months, which is more than all of 2024 and 2025 put together.

Predictably, the company says that agentic AI is behind the dizzying rise in the amount of code, and the need for it to both upgrade and rethink its technological infrastructure. By last September, over 40% of pull requests were started by an agent, up from 17% in March 2026. Moreover, repositories that kick off using AI generally see eight times the activity compared to handcrafted operations.

All this code activity puts disproportionate strain on CI/CD pipelines. In modern development, whenever someone commits code or merges a pending pull request, many actions are triggered: for example, checking for leftover "to-do" items, visually arranging code for consistency (linting), and running code and integration tests to check that the new code doesn't break existing functionality within the application and works correctly with associated programs, services, and APIs.

Triggering all of the above for each commit and/or pull request means that GitHub's hardware requirements have ballooned, now requiring massive amounts of computing power and I/O. The rise was so intense that the firm even had to reengineer its backend to try to decouple storage from compute as much as possible. Given how quickly demand is rising, GitHub is likely to have to upgrade sooner rather than later.

Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds.

Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.

Bruno Ferreira is a contributing writer for Tom's Hardware. He has decades of experience with PC hardware and assorted sundries, alongside a career as a developer. He's obsessed with detail and has a tendency to ramble on the topics he loves. When not doing that, he's usually playing games, or at live music shows and festivals.