Most of us learned to think about cloud workloads in terms of one bill: the dollar cost. There is a second bill arriving, and it is denominated in carbon. GreenOps is the practice of treating that carbon cost as a real, measurable, optimizable number, the same way FinOps treats dollars. And here is the part that makes it worth an engineer's attention rather than a compliance slide: most of the moves that cut carbon also cut cost. GreenOps is largely FinOps wearing a different hat.
Let me lay out what carbon-aware infrastructure actually means in practice, without the hand-waving.
The two levers: when you run, and where you run
Carbon-aware computing has exactly two knobs, and both are things schedulers already understand.
Time-shifting (when). The carbon intensity of the electricity grid changes hour to hour. Midday with lots of solar on the grid is cleaner than a still evening running on gas peakers. A batch job that does not care whether it runs at 2pm or 2am can be scheduled for the greenest window. This is the same idea as running non-prod off-hours for cost, pointed at a different signal.
