Welp, lets see if a new PSU fixes my 7900GT issues.
I've RMA'd 3 cards back to EVGA now and lastnight, the 2nd replacement card started artifacting in EQ2 and it thus began the same problems all over again that thousands of people are having with this series of card.
I called EVGA tech support and they wanted to know the usual questions. They said it may be a PSU issue, but I did and still do disagree that a 26amp on +12 (with voltage at 12v measured) PSU is my problem. So just to take that out of the mix, I bought a Silverstone 560 with 38amps on the +12, so that if at some point I can run 2 cards in SLI again (since these 7900's seem to be pieces of horse***t) there won't be any questions about power issues.
As you can tell, I'm pretty damn fed up with this issue. I could understand it on the OC'd cards, but were talking about cards that are at FACTORY settings!
Anyhow, all the normal Nvidia manufacturers are having to deal with this problems as RMA's have went up as high as 5% just over these cards.
Sledge, if yours are not giving you any trouble yet then you are one of the lucky few.
I called EVGA tech support and they wanted to know the usual questions. They said it may be a PSU issue, but I did and still do disagree that a 26amp on +12 (with voltage at 12v measured) PSU is my problem. So just to take that out of the mix, I bought a Silverstone 560 with 38amps on the +12, so that if at some point I can run 2 cards in SLI again (since these 7900's seem to be pieces of horse***t) there won't be any questions about power issues.
As you can tell, I'm pretty damn fed up with this issue. I could understand it on the OC'd cards, but were talking about cards that are at FACTORY settings!
Anyhow, all the normal Nvidia manufacturers are having to deal with this problems as RMA's have went up as high as 5% just over these cards.
Sledge, if yours are not giving you any trouble yet then you are one of the lucky few.
0
Comments
12v * 38amperes = 456watts.
You say its a 560 watt psu? how much current is on the other rails. it looks like next to none, shoot. unless it just plain isnt going to put out the rated specs.
Which is so common it isn't even funny in commercial PSUs.
This is why there are so many reports of eVGA 7900GT OC'ed cards going south in the various forums out there. XFX is also suffering this problem.
http://www.newegg.com/Product/Product.asp?Item=N82E16817163111
Madmat, you are corect with the fact that they have factory OC'd these cards to high. Unfortunately at this point I can't even down clock the card to stability, and its happening across both GT and GTX lines.
I'm very tempted to get 2 7800 512's to replace these with.
Silverstone offers some very nicely built PSU's, I know that they don't build them but they're built by a top notch supplier. Silverstones are either built by Enhance or Etasis.
Please read the article; it just might help you out.
PSU is less then a year old currently. Case is a highly modified Chieftek server full tower. Its huge =), I wanted something I could move air around in easily.
Mud, to answer your reply I have already RMA'd 2 cards back to EVGA and the very first RMA was to newegg. Looks like a 4th is going back to EVGA this week sometime also. I have the card I recieved from the last RMA sitting on my desk at home. I haven't even put it in yet as i'm waiting on the new PSU. I saw that article when it hit last week =D
I guess you are just having bad luck or you have a defective mobo sending to much power to the GPU.
What series of EVGA cards are you using? This problem is isolated more with the 563AX or CO's then the KO's or Signature Series.
Model #'s
eVGA 256-P2-N564
eVGA 256-P2-N567
eVGA 256-P2-N569
eVGA 512-P2-N575-AX
and for the XFX series
Model #'s
PV-T71G-UCE7
PV-T71G-UCF7
PV-T71G-UDE7
XFX PV-T71F-YDL9
XFX PV-T71F-YDE9
I'm having to rma one of my XFX 7900gt's and few other people on another forum are reporting issues with them.
520M
Got this from a post on Hexus.net
Now obviously this may only work XFX cards but it's worth giving it go.
Link to thread
I've used coolbits, powerstrip, rivatuner, atitool and just about every prog to test and try to fix the issue.
No juice so far.
Have you tired auto detect?
Mack as far as I see tell they follow the reference design so it would suggest to me a dodgy bios revision but I'll bow to your better knowledge of these things. I suppose the memory chips may be a different revision or something.
Fails