If you've been reading this for a while, you'll know I had a bit of trouble getting my new system working. Eventually I did, with a new GTX670. Everything was entirely fine - as in, it couldn't have worked so well for so long without being so. I must've put in 100 hours GPU strain-time or so over a couple of months, and, while I had a couple of crashes, we all know, sa PC users, that doesn't mean a great deal. There was nothing to indicate it was GPU related other than fans going mental as it blackscreened on me. it only did this once or twice, and would always just load again. I checked temperatures and everything said it was fine - both the system temperature generally and the gpu control panel thing. So.. then it had a crash from which it appears to have died. If I turn it on at all now, the system seems to be booting fine but the screen's dead and the GPU fans spin up to crazy and just stay there. I'm assuming it's fried, somehow. Any ideas? Sounds like an RMA to me, this shouldn't have happened. Anyone know if it'll need to go back to ebuyer, where I bought it, or.. I think it's EVGA - direct?
Incidentally, it had been chugging alone ostensibly fine with Far Cry 3, Dishonored, Planetside 2, all at max settings. Yet when it had it's final moment, I was in Crusader Kings II. Go figure...
Always worth checking it in another computer before you ship it off for RMA. If possible. Not sure what the RMA procedure is in the UK, but around here the (r)etailer is obligated to deal with it (so there's no reason to send it to the manufacturer).
In the UK, if you send it the manufacturer, you surrender your rights under UK consumer law and put yourself entirely at the mercy of the 'warranty terms' - for which you have little legal comeback if they choose to mess you around.
That's not to say it's not always a good idea to deal with the manufacturer - some can be very good in this respect but you are giving-up any rights your have by doing so - so it's worth trying the 'proper' way first I reckon
It's pretty much essential to test the card in another system, simply because it rules-out your mobo or PSU as the source of the problem.
If you can't do that, you run the risk of wasting time and money returning it for them to find it's AOK
Sure. I'll run around with this stuff again (had to troubleshoot some eventual non-issues putting it together), but I can't see it being anything else. If I put my old card in it and it's fine, that's got to be the issue. The way it's acting seems to be way off. It does seem pretty odd, though, to die after starting to get established in life. Why always the young!
Just to be clear about this - the things which will cause issues with GPUs are
a - the GPU
b - your motherboard
c - your power supply
That is both actual issues can be caused and actual damage can be caused by those things - and not, necessarily in that order.
So it's possible your issue is that your motherboard or PSU is frying your GPU and that a replacement gpu will go the same way. Trying the card in another PC is sensibly ruling out an issue with those parts now but doesn't mean they won't kill another card anyway.
So long as the gpu is less than 6 months old, you should get a replacement - if it's over 6 months old you might have to argue the toss (in consumer law, you'd have to prove the card is defective and that you didn't kill it through abuse/neglect/other faulty components)
Are you absolutely sure your PSU isn't the issue here because your GPU is the biggest power used by FAR - unless it's a high-spec and reasonably new PSU I'd probably be thinking "replacement" myself.
Not necessarily. For example, if the GTX 670 is overloading the power supply and your old card uses less power.
Originally Posted by TheRed
Well - if you remember, we talked about all this. I had suspected that the original difficulties may be PSU related, so I got rid of the 550W one, which had, as it turned out, not been an issue, and got a 750W one - so that really shouldn't be it. Of course I will try it in this old one I'm on now (again!) That PSU was new towards the end of December, too, brand spanking. It's evident I'm no expert, but it is fucking annoying that something can ostensibly work 100% for a couple of months and then just asy "No ta".
Light bulbs work until they fail.
Someone last week told me that they always pit their car at trackdays before their last intended lap - when I asked why they said "it's always on the last lap when you crash" to which I said "yeeeeesssss, it usually is!" :)
Quite so. And string is as long as a piece...!
Still annoying. I'm fairly sure - and perhaps you can reassure me here - that I hadn't done anything 'wrong' if it worked pretty damn beautifully for a couple of months. These things happen I guess, but I had a right hassle getting the thing going in the first place, due to one faulty card and one odd issue with DV output.
Only in Need for Speed games with rubber banding that forces you off the road. Oh, and overdoing the "last lap" mentality IRL. :D
Originally Posted by trjp
Oh, as to Ebuyer they have been good to me in the past with RMAs. Not sure how their current service is as not needed to use it in a while (though they answer questions quickly). If the card is dead, I say RMA it. If you changed/overclocked/broke it yourself though, it's not their responsibility to offer you a refund.
When I get a moment, I'll try it in the other machine etc. I think an RMA will be needed, though - and I certainly haven't touched it one little bit. No over-clocking, nothing.
Originally Posted by TechnicalBen