Computing costs and local AI
Whether renting out a GPU or server pays, running AI models on hardware you already own, and cutting what your build pipeline costs.
The hosting guides explain how power, platform fees and utilization affect rental income. They are separate from the local AI experiments, which measure inference on hardware we already own.
Engineering notes cover those experiments and our GitHub Actions cost audit. They report the setup, measurements and limits so you can judge whether a result applies to your own hardware or workflow.
7 articles by Steven Michael Barber.
When trimming CI was not enough: moving the jobs to a desktop we already owned
Engineering note
Why more pipeline trims could not reach our GitHub Actions target, four paths compared on cash, and the two-hour setup of a self-hosted runner on a desktop.
How we audited GitHub Actions cost and run time
Engineering note
How we audited repeated CI work, separated waiting time from runner minutes and chose simpler checks with measurable trade-offs, with a worked example.
We split one AI model across two home computers. In our tests, it got slower
Engineering note
We split Llama 3.1 8B and 3.3 70B across a desktop and 2017 iMac over Wi-Fi. All splits generated text more slowly; bottleneck causes need more tests.
Our 2017 iMac couldn't run Ollama on macOS. On Linux it reached 70% of a modern desktop
Engineering note
In our October 2026 test, Ollama needed macOS 14; on Ubuntu, a 2017 iMac ran 3B and 8B models at about 70% of our Dell’s speed.
We built a private AI server without a dedicated GPU. Here's what our first benchmarks showed
Engineering note
Three local models on a Windows desktop with no dedicated graphics card, run with Ollama: 15.67, 7.67 and 4.26 tokens per second, and how we measured.
Utilization: the number that decides whether a rented GPU pays
Practical guide
What the share of rented hours does to one GPU's net income and payback, the three utilization thresholds that matter, and the payout trade-off.
Is renting out a spare GPU or CPU worth it?
Practical guide
How power draw, electricity price, utilization, payout and platform fees combine into net income per hour and a break-even date for a rented GPU or server.