Because of the enormous number of components, special configurations, drivers, and operating systems, this question becomes almost automatically one of the most difficult to answer. However, with the right tools it is possible to reveal some very important details. For example: what role does hardware play in that performance loss? Thanks to an excellent opportunity, LinusTechTips compared a GeForce GTX 480 that was used hard for more than half a decade with one straight out of the box, untouched.
Do you feel your computer is "dragging" a bit? Don't worry, we've all been there. Over the years we've learned to accept it as inevitable, but deep down that doubt always lingers...
Why Does a PC Lose Performance?
Even with marathon and almost obsessive maintenance sessions, the user is fighting a losing war, and those who lack the knowledge or patience to try to correct this situation decide to cut their losses and buy new hardware. Now, how responsible is the hardware? Can we talk about a "sustained degradation" of the components?
That idea arises especially in the case of graphics cards, when they are sent to major specialized media outlets to publish their benchmarking results. The fact that some cards are sent back and then transferred to new portals that present additional benchmarks suggests the possibility of "wear and tear," that is, that the card does not perform as well at the fifth or sixth performance study as it did at the first, due to the punishment it suffered.
The most logical way to verify this is to compare a piece of hardware widely used for several years with an identical one that has never been installed in a computer. With that in mind, LinusTechTips managed to get its claws on a GeForce GTX 480 NIB (new-in-box) that the people at EVGA sent. The GTX 480 was released in March 2010, so a margin of more than six years should be more than enough to determine whether that sustained degradation of hardware exists or not.
The "new" GTX 480 faced a twin reference card used without mercy in benchmarks and games. Crysis 3, 3DMark 11, Batman: Arkham Origins, and GTA V were the tools used in the comparison of the two cards, and the conclusion is... that there is no degradation.
Use Doesn't Degrade Performance
That's right: despite the huge difference in activity, and variables like heat exposure and the stress of multiple physical installations, the two GTX 480s ended up virtually tied in the benchmarks, with differences of one frame per second that fall within the margin of error. If the manufacturer does its job (by which I mean "not cheating," replacing components with lower-cost ones), the most critical point of a graphics card is its cooling system, since it has moving parts.
After a sufficiently long amount of time, failures will be inevitable for both cards, but they will be spontaneous and immediate failures (freezes, artifacts, or rendering errors), not something that can be measured in a benchmark. The story is different for hard drives and solid-state storage, whose degradation is more evident thanks to the S.M.A.R.T. system.
However, I share the conclusion reached by the video: the main responsible for performance loss in a computer tends to be software, whether it's bloatware, bad drivers, malware, too many background processes, or updates that weren't properly optimized. In summary: Less is more.
