SHARE
Can your device degrade over time from regular use?

Should we buy second-hand gadgets?

I think it’s pretty safe to say that anyone who’s purchased an electric “DooDa” of some sort, has experienced performance degradation over time. How’s that iPhone 4 running for you these days?

Why does your device or hardware slow down?

So what exactly causes your devices to go from hero to zero over a span of just a few years? These Genius Tech Tips will reveal.

This investigation into performance loss was actually inspired by a conversation that I read over on Reddit about the Radeon RX 480 graphics card announcement. One of the commenters suggested that AMDs practice of loaning graphics cards to reviewers, then sending them over to another reviewer could hurt their benchmark scores, due to the cards becoming worn out.

gfx-card-performance-conversation-on-reddit
Reddit Conversation about Video Card Degredation

Reddit members discuss about a video card degrading over time. The back and forth discussion that followed is something I have seen a fair number of times. Usually In the context of, whether or not it’s safe to buy used computer parts, especially cards that have been used for Crypto Currency Mining, but this was a theory I’ve never actually seen tested.

So how do we test this theory?

Now the problem is that new hardware is impossible to go forward in time and make very old, so no investigation into a GTX 1080 performance degradation over time can really be done. Also old hardware is very difficult to find still in its brand new condition, that is to say unless you’ve got friends in “low” places, like EVGAs warehouse for example. Where just such a thing was found by my buddy Adam.

A brand new, sealed GTX 480. Vedeo card performance test
A brand new, sealed Nvidia GTX 480. Video card!

So I managed to get my hands on a brand spanking new, still sealed reference Nvidia GTX 480. A 6 year old graphics card! I also managed to get hold of a reference GTX 480 that was extensively used for benchmarking. I moved the card over to my personal rig, then used it for benchmarking some more. Also Plenty of rough genius tech tips handling in between.So if there’s any card out there that could have degraded due to hard use, mistreatment or heat? This GTX 480 I had, was it.

So I unboxed the new 480 in an attempt to compare, and see if the used one had in fact degraded over time after subject to regular use. It was crazy to see such an old product in brand new packaging.

Unboxing the New Radeon GTX 480.
Unboxing the New Nvidia GTX 480.

Now I did think they may have just sent me an old card, but in a new box, however you can tell if the card is old. With the GTX 480 the surface of the graphics card stains when you touch it. So I could tell this new card was actually, legitimately brand new and that it had indeed been in the box for somewhere between 4 and 6 years. It was a pristine condition graphics card.

So what happened after testing? The big reveal!

For my test bench to avoid any hardware bottle necks, I used a very different class of hardware than what was available in the GTX 480 era. A core i7 5960x 8 core extreme edition, 128 gig of DDR 4 ram and all of that running on an Asus x99 deluxe 2 motherboard.

I began this test with first a look at temperature, raw clock speeds and compared it to what is advertised for the card. If the card is actually throttling down whilst under heavy load, then any discrepancy here could have saved me a lot of trouble running any other tests.

Radeon GTX 480 Specifications.
Nvidia GTX 480 Specifications.

For this test I used a Crysis 3 Skybox thermal test, which is both real-world and puts the card under 100% load, spiking it up to almost 90°c on the GPU core. Even on an open air test bench, with a fan blowing down on to the video card. To run these temperature an clock tests I used the faithful MSI Afterburner.

A reminder to the “fanboys” out there. life is regrettably not as simple as “AMD cards run hot” or “Invidia has better drivers”. It varies significantly from product to product.

Using Crysis 3 to determine GTX480 temperature and core clock.
Using Crysis 3 to determine GTX480 temperature and core clock.

Anyway, speaking of “varying significantly”, our two cards didn’t! Both cards behaved identically after 10 minutes of Crysis 3, with the brand new card acquiring 1% more fan speed for the same 701 MHz core clock speed, and 90°c operating temperature.

Old and New Radeon 480 GTX Temperature Results
Old and New Nvidia 480 GTX Temperature Results

3D Mark Tests

3D mark usually has some variation from run to run, but both of my 480s got s*#t on, just about equally. Futuremark compared our rig to a gaming laptop from 2013 despite the beefy CPU we used.

Well Used Radeon GTX 480 3D Mark Benchtest Result.
Well Used Nvidia GTX 480 3DMark Result.
performance-degradation-is-it-real-1-0051815
Brand New Nvidia GTX 480 3DMark Result.

What about the GTX 480 running some real games?

Crysis 3 at high detail and textures shows us that the GTX 480 is still a reasonably competent gaming card, whether being brand new or indeed very old. I guess this test is also a testament to the reliability of our Crysis 3 benchmark run, because both cards came in pretty damn near the same.

Nvidia GTX 480 Crysis 3 Game Performance
Nvidia GTX 480 Crysis 3 Game Performance

The canned benchmarks in Batman Arkham Origins and Grand Theft Auto 5 results tell a similar story.

Nvidia GTX 480 Batman Arkham Origins Game Performance
Nvidia GTX 480 Batman Arkham Origins Game Performance
Nvidia GTX 480 GTA 5 Game Performance
Nvidia GTX 480 GTA 5 Game Performance

Leading us finally, anti-climatically to the conclusion.

Gum Flavor Potency Over Time (Not Overclocked)
Gum Flavor Potency Over Time (Not Overclocked)

Other than maybe a cards cooling fans which will whine or stop out right when they fail, video cards like most computer hardware have no moving parts. Video cards do not generally suffer from gradually declining performance from regular use over an extended period of time. In the same way that, let’s say a stick of chewing gum would.

This however DOES NOT mean that Video cards don’t ware out. Electron migration (the momentum of electrons) eventually cause atoms to move, that can cause circuits to fail and this is accelerated by heat, or by over clocking your graphics card. The way that graphics cards fail are not generally gradual or predictable.

It’s more like the way your house roof would fail. It totally works until such time it totally doesn’t work, and you get water on your head giving you a fairly good idea that it’s time for a new one!

Hard drives and solid state drives will degrade over time!

So if your iPhone or your laptop, like a video card doesn’t have any moving parts?

Then based on our testing.

The reason why your device or hardware slows down (and if its storage device is still running well) is more likely to do with software, bloated applications, anti-malware packages or indeed malware (pick your poison right?) running in the background. Operating system updates that are designed with more overhead for newer, faster hardware is also among the most common culprits of slowing down your device.

So If your looking for advice on buying a second hand device, we recommend looking through the details as to how well the device has been looked after, but as for how old the device is? This generally shouldn’t be a problem with graphics cards or other devices without storage hardware build in.


LEAVE A REPLY