Does the warranty cover 'over-use'?

Nowadays, most warranties for video cards are lifetime. But with all these powerful GPUs capable of doing so much more (ie Folding), do you think the warranty would cover its failure if you were running it 24/7 full load?

Many seem to speculate that video card lifetime is far less than CPU lifetime...So if I were to use my GPU to fold 24/7, does that legitimately fall under their normal 'wear and tear' if I never modify anything or even overclock?

Would the signs of 'over-use' failure look very similar to a overclock failure?

Any failure of the Video Card during normal operation (ie. no overclocks, no tamering, etc) should be covered by warranty.

Folding doesnt really use GPU....anyway.

Warrantys are based off 24/7 use for the entire period, not periodical use. Also, if you keep a card going full-ball every hour of every day without turning it off, its actually probably better for it because it doesnt get exposed to varying conditions which may fatigue it.

Hammer your GPU all you want

So changing temperatures is more detrimental to a GPU with stock cooling than a constant high temp?

Folding is currently developing code to run specifically on GPUs (Gromacs). Apparently modern day GPUs are very well suited for these types of calculations. There was a benchmark done on unoptimized code which showed a Radeon x1800XT processing almost 2.5x faster than a 3.0Ghz P4.

In light of that observation of varying tempertures, is it better to leave your PC on for long periods of time and reboot it occasionaly or shut it down when your not using it, say every night when you go to bed. I wonder.

ZOMGWTFBBQ said:

Folding doesnt really use GPU....anyway.

Warrantys are based off 24/7 use for the entire period, not periodical use. Also, if you keep a card going full-ball every hour of every day without turning it off, its actually probably better for it because it doesnt get exposed to varying conditions which may fatigue it.

Hammer your GPU all you want

Click to expand...

I dont think that is true. Something with moving parts (i.e. a hard drive) that might be true, because the hardest the component has to work is when its spooling up or searching for something. On the other hand, a VGA works hardest when it is going full throttle. Idle or off require the least amount of work from the video card, while when its in full use, it is the hardest for it. Running hardest all the time will make it burn out more quickly than changing conditions.

devman said:

In light of that observation of varying tempertures, is it better to leave your PC on for long periods of time and reboot it occasionaly or shut it down when your not using it, say every night when you go to bed. I wonder.

Click to expand...

The longer you run anything, the less life it will have. Almost every electronic device is rated in hours so shutting it down would elongate the life. Although computers are rated for extreme numbers of hours before failing.

HotkeyCC said:

The longer you run anything, the less life it will have. Almost every electronic device is rated in hours so shutting it down would elongate the life. Although computers are rated for extreme numbers of hours before failing.

Click to expand...

Just remember that each time you start something up, it gets kinda bombarded with elecrons. Constantly shutting down and restarting a PC wll kill it pretty quickly, too.

My PC runs 24/7. The oldest component to be replaced by "newer" tech rather than failing is my SB Live! 5.1. Bought in Jan 2001, retired Jan 2006. It's seen 3 different motherboards and 3 different cases. I got my money's worth for that card.

I've done some preventitive maintenance on my BFG 6800 GT, I've put on a NV Silencer Fan (I wasn't a fan of the 2 fan/LED one) and I've added a support bracket on the side. The stock fan weighs MORE than the silencer and caused the card to "droop". The silencer is lighter but still the card drooped in the slot. I expect this card to last as long as my gf3 which was replaced in this system around the time the 6800 gt became readilly avialable and was in my system as long as the Live! 5.1.

The usage life of most electrical components is tested around the average 10 year mark. It's done through a calculation and excessive stress testing and taking into account the parts used (capacitors, resistors, pcb lifespans etc).

Some parts last longer, some shorter, it depends on the conditions its faced with.

One thing I do not like about new cards is the sheer size of them. They are simply too large, then you add in that the heatsinks are massive and weigh down the beasts. Physical stress is probably not taken into account these days...

I'd seriously take a look at your video cards and see just how much they are "flexing" in your systems...

I assume the idea behind running it for long periods of time rather than shutting it down and turning it back on in short intervals is based on the stress caused by the expansion of different parts at different rates.

But is this really a problem for computer components?

I dont think so. When you start a computer(most computers not including like a PDA processor that changes voltages as is required by the software) it is subject to the same amount of electricity (Within .05V or so) as it will be the entire time its running. Shutting it off and turning on really doesnt make any changes to the device itself. I think just about anything that is metal will expand slightly when heated up, but I dont think it will reduce the life of the card, since the engineers probably understand this fact far better than we do and take steps to accomodate it.

i only turn mine off during lightning storms :O

"Normal" use is 24/7/365 for many people, and I have yet to have a RMA denied because of this level of use.

My latest return was directly to 3Com, for a 6 year old PCI 10/100 ethernet card that wasn't working too well. It had recently been in a computer that had been on continually for the last 4 years, and in another computer that was continually on for 2 years before that.

No problems here, and I got a brand new one from them. Nice folks.

Before that, I returned an ATi Radeon 9700 Pro card directly to ATi, and it had been in a machine that was continually on for the last 2.5 years. No problems here, either.

I dont think its possible to tell how frequently it's on anyways. Not like there's an odometer on the video card

actually, a lot of hardware devices do have them, and it wouldn't shock me at all if there is one burried somewhere in the ATI cards, I know I have a creative video card (PCI) around here somewhere that had at least an uptimer (made me laugh when I saw the card with an uptime count)


...Nice high-quality S-vid output on that card...

(Might have to break out my old P2-233 and set it up as a SNES/MAME emulator on the big screen )

’m‚³‚ñ said:

Just remember that each time you start something up, it gets kinda bombarded with elecrons. Constantly shutting down and restarting a PC wll kill it pretty quickly, too.

Click to expand...

Does this still apply? Modern mobos, especially the ones aimed at modders, boast multiple regulators built into the board (I think my last mobo had six next to the CPU).

They've also made self parking harddrives for that reason too. In the old ones, if there was a surge during startup, whatever data was below the drive head when it turns on goes pfffft

All that aside, is there some sort of instructions or specifications included with your videocard specifying how much airflow you need? If you run games all day, in a closed environment, with a sealed case in the tropics, it can get pretty hot. I'm serious, i live in the tropics, it's early morning in january and i'm already sweating!

Sly said:

Does this still apply? Modern mobos, especially the ones aimed at modders, boast multiple regulators built into the board (I think my last mobo had six next to the CPU).

They've also made self parking harddrives for that reason too. In the old ones, if there was a surge during startup, whatever data was below the drive head when it turns on goes pfffft

All that aside, is there some sort of instructions or specifications included with your videocard specifying how much airflow you need? If you run games all day, in a closed environment, with a sealed case in the tropics, it can get pretty hot. I'm serious, i live in the tropics, it's early morning in january and i'm already sweating!

Click to expand...

I'm not sure, actually. I do know that standard light bulbs have this problem, turning the lighgt on causes a huge jump in power draw. Though, energy saving lights have changed this. IIRC, it has to do with heat, until your equipment gets to spec operating temp, it will require a little bit more power, I think.

I can imagine, sometimes it gets up to 120 F here during the day, often when it doesn, it hardly cools down at night. Usually, when it's that hot, I simply leave my PC's off unless I need to surf.