Manufacturer battery life claims are typically benchmarked against something like local video playback at a fixed, moderate screen brightness — a workload that keeps the CPU at a low, steady, largely idle load for the entire test, which is about as gentle a real-world task as a laptop performs.

Kicking off a full build asks something completely different of the hardware: it pins most or all CPU cores at high utilization simultaneously, often for sustained minutes at a time rather than the brief bursts most everyday tasks produce — a fundamentally heavier and less power-efficient workload than the steady, low-load task the marketed number was measured against.

The direct consequence: sustained high CPU utilization draws meaningfully more power than the video-playback benchmark does, which is why a laptop rated for "18 hours" can visibly drop battery percentage in real time during an actual build — the rated number was never describing that kind of load in the first place.

A second, related effect is worth naming: sustained heavy compute on a thin or fanless laptop design can also trigger thermal throttling, where the chip deliberately slows itself down to stay within its temperature limits. That means a long build on a thin laptop can end up both draining the battery faster and running slower than the same build on a thicker, better-cooled machine, even with an identical chip inside both.

Plugging in during a heavy build sidesteps the problem, obviously, but it's worth knowing the shape of the gap before you're caught without an outlet: a laptop that comfortably lasts a full workday on light tasks can burn through several hours of that same battery in well under an hour of sustained, all-core compiling — a real constraint for anyone doing genuine development work away from a desk, and a good reason to check a laptop's sustained (not peak) performance rating before relying on it for that kind of work untethered.