You can also run the system like a server and set it up and then SSH into the system for control, and not have a video output, meaning less work for the card.
That's nothing — compared to a typical mining load, a video output doesn't add any significant "wear" on the card at all.
The main killers for graphics cards are swings in voltage/temperature (mining on, then mining off, then back on etc — when the elements on the card and the gpu itself are heated up to 70-100 degrees, and then rapidly cooled down to 30-35 and vice versa — the more cycles like this happen, the more likely card to fail) and abnormally high temperatures in general — which usually happens with the parts of the card that aren't being monitored: among all those people that run their "gpus" at 70-80 degrees (= actual gpu temp), some not so well designed cards might have their vrms at 100+ degrees, and while it's ok temps for vrm's themselves (they're usually good up to something like 150 degrees), the elements around vrm might not like that heat and it'll shorten their lifespan).