New graphics card problem

[Jakus]

I'm using MSi mobo and MSi GTX 660 Ti and they are great.

There really does seem to be an issue with the MSi HD 7950s though.

There seems to be a bunch of guys around here into these 7950's and an inordinate amount of faulty cards....suggest someone starts making note of what memory chips they are running on these duff cards.

I helped a guy the other day who was on his 3rd 7850, think they were MSI also

They may just have some out of spec surface mount resistor that's lowering the memory vrm, I know some can software adjust the memory voltage but those type of tweakers just report expected voltages not measured....anyway your average joe should not have to be fault finding faulty new cards !

 

[Geo-Gamer]

My Card will be here on Monday/Tuesday dont scare me! That MSI might be faulty!

It looks like its all down to luck! Thankfully I ordered two msi 7950 tf3 and one of them being just over a week ago and they work 100% correctly .

Fingers crossed for you mate!

 

[Jakus]

Just had another thought, Rusty said don't buy the MSI 7950 on special and Gibbo said They got a new batch and all the old batch was sold !! well, that will teach OCUK for buying all the allocation up

I still have a weird feeling these cards are EOL, no solid data but a gut feeling !

 

[Jakus]

It looks like its all down to luck! Thankfully I ordered two msi 7950 tf3 and one of them being just over a week ago and they work 100% correctly .

Fingers crossed for you mate!

Listen, It would be impossible for manufactures to test the cards properly all you have to do, and most do this anyway without realising Is run the card through a bit of gentle load for 15 mins (installing drivers) then bench it and monitor the temps...then make a fan curve etc.

Make sure it will overclock at least 50mhz 100 is better also the memory even if you will never overclock it's no good if the card has no headroom reliability wise.

So after a few hours if you have no problems, that card should now last you for as long as you want to keep it, providing you keep the temps down and always let it cool down after hard use for 10 mins.

It's the nature of electronics that if it's ok for the first 1% of it's life it will be fine until 85/90% of it's life expectancy, providing excess heat and thermal shock are avoided.

I'm not a gambling man but being realistic I would test those cards thoroughly individually with the expectations they are from the faulty batch
but don't worry about it, what will be will be...and I hope you lucky

 

[Geo-Gamer]

Listen, It would be impossible for manufactures to test the cards properly all you have to do, and most do this anyway without realising Is run the card through a bit of gentle load for 15 mins (installing drivers) then bench it and monitor the temps...then make a fan curve etc.

Make sure it will overclock at least 50mhz 100 is better also the memory even if you will never overclock it's no good if the card has no headroom reliability wise.

So after a few hours if you have no problems, that card should now last you for as long as you want to keep it, providing you keep the temps down and always let it cool down after hard use for 10 mins.

It's the nature of electronics that if it's ok for the first 1% of it's life it will be fine until 85/90% of it's life expectancy, providing excess heat and thermal shock are avoided.

I'm not a gambling man but being realistic I would test those cards thoroughly individually with the expectations they are from the faulty batch
but don't worry about it, what will be will be...and I hope you lucky

Thanks for the advice mate

So far so good it seems. Got it overclocked 1150/1700 @ 1.14 and played all my games with no hiccups xD! Until I switched to bf3 and saw very slight annoying artifacts so I then upped the voltage from 1143 to 1156 and it's running like a dream on all my games. Got the fan set at 60% and I'm very happy as the gpu max temp is 61c and vrm max temp being 69c. So far all seems brilliant and very happy with the card

Quick question mate what will you recommend the max safe temps for the vrm temps? Thanks.

 

[Jakus]

Thanks for the advice mate

So far so good it seems. Got it overclocked 1150/1700 @ 1.14 and played all my games with no hiccups xD! Until I switched to bf3 and saw very slight annoying artifacts so I then upped the voltage from 1143 to 1156 and it's running like a dream on all my games. Got the fan set at 60% and I'm very happy as the gpu max temp is 61c and vrm max temp being 69c. So far all seems brilliant and very happy with the card

Quick question mate what will you recommend the max safe temps for the vrm temps? Thanks.

All electronics is the same cooler it runs the longer it will last, personally I think 70 c is too hot 60c is a nice practical temp for things like GFX cards with air cooling. Same goes obviously for the vrm's. For 24/7 I would take a little off the clock and voltage and see what I really lost performance wise but the fact you have an eye on it , I don't think you will have any problems.. some people unknowingly get things like vrms up to 90+

Great temps on the core though

 

[x22dan]

Listen, It would be impossible for manufactures to test the cards properly all you have to do, and most do this anyway without realising Is run the card through a bit of gentle load for 15 mins (installing drivers) then bench it and monitor the temps...then make a fan curve etc.

Make sure it will overclock at least 50mhz 100 is better also the memory even if you will never overclock it's no good if the card has no headroom reliability wise.

So after a few hours if you have no problems, that card should now last you for as long as you want to keep it, providing you keep the temps down and always let it cool down after hard use for 10 mins.

It's the nature of electronics that if it's ok for the first 1% of it's life it will be fine until 85/90% of it's life expectancy, providing excess heat and thermal shock are avoided.

I'm not a gambling man but being realistic I would test those cards thoroughly individually with the expectations they are from the faulty batch
but don't worry about it, what will be will be...and I hope you lucky

once again i tried uninstalling, reinstalling benchmarking ect ect
and still im getting artifacts all over the screen even while my computer is idle.
Reverting back to my old card until OcUK contacts me back

 

[Jakus]

once again i tried uninstalling, reinstalling benchmarking ect ect
and still im getting artifacts all over the screen even while my computer is idle.
Reverting back to my old card until OcUK contacts me back

As you picked it up from the shop if I was you I would go straight back and ask for a replacement.

You are obviously logical and sensible and have proved the card faulty but the only issue is you are not covered by the DSR so technically it's all on you to return the card through their system which I imagine is not smart enough to expedite genuine faulty products.

In your case if it's hard to get there, I would telephone a real person and buy another over the phone and ask them to pick the other one up at delivery.
remind them, while not their fault there has been a run of similar problems

may be worth posting here also....
http://forums.overclockers.co.uk/forumdisplay.php?f=116

 

[wazza300]

soundslike gpu ram issue if its artifacting? its running at idle clocks in desktop? try downclocking the gpu memory and retest