Can a power station run a PC?

13.2 W while it is running, and 2.3–37 W across the models in this class, with 4,600 W demanded at the instant it starts. Those are different questions and most pages answer only the first. The running figure decides whether a machine can hold it; the start-up figure decides whether it can get it going at all, and it is usually the one that fails. The start-up figure here is the manufacturer’s own..

For a desktop PC, the BLUETTI Elite 30 V2 (US 120 V) is the cheapest machine in our library that runs it, at 600 W continuous and $239 against a running draw of 13.2 W. That is a verdict computed from this load’s running watts, its start-up demand and the inverter waveform — not a rule of thumb about doubling the wattage. 19 of 26 machines clear it, 7 are capped below green rather than shown to run it — this load is documented by its maker as needing a pure sine wave, and we hold no inverter waveform for those builds, and a check that could not be run is not a check that passed.

It depends entirely on the machine, which is the part most pages skip. Against a desktop PC at 13.2 W of running draw, our figures put the BLUETTI Elite 30 V2 at about 11 h, and the Jackery Explorer 300 Plus at about 9.2 h. Those hours come from usable-capacity figures derived from each manufacturer’s own published usable-capacity coefficients — depth of discharge and inverter efficiency where a brand states them separately, with each machine’s idle draw in the denominator — not from dividing a nameplate watt-hour number by a label wattage. Most machines in the list carry no hours at all: we hold nothing for those builds that makes a usable-capacity figure defensible, so those rows give an energy rate instead of a number we would have had to invent.

Less than the number suggests, and this site says so rather than quietly using it. The start-up figure is published by the manufacturer for this device, so it describes the actual hardware rather than a category. The measured spike is 4,600 W, but it lasts a few line cycles — a capacitor charging, not a motor starting — while a station's surge rating describes a second-scale event. They are not the same quantity, so the spike is reported here and deliberately not held against any verdict above. A site that compared the two would fail this load against machines that run it perfectly well.

Yes on our sources, and unusually this is a manufacturer statement rather than an internet one. The requirement here is inherited rather than device-specific, and the reasoning IS the claim, so the page states it: an 80 PLUS-certified PC power supply has to reach a power factor of 0.9 or better, which in practice means active power-factor correction. CyberPower’s own compatibility note warns that a simulated-sine output produces a momentary zero-output gap at each phase change, and that an active-PFC supply fed from one can shut down hard or fail outright. That is a manufacturer warning about a class of power supply, not a measurement of any one computer, which is why this field’s confidence is capped below the specs around it. If your machine’s supply is not 80 PLUS certified the chain above does not reach it — check the supply, not the computer. Where we hold no inverter waveform for a station build, this site refuses the green light rather than assuming one — which is why 7 of the 26 machines above sit at amber.