7900GTO Failure Rate Poll

How soon did your 7900GTO start artifacting / failing?

  • Less than 2 weeks

    Votes: 9 12.9%
  • 2 - 4 weeks

    Votes: 6 8.6%
  • 4 - 6 weeks

    Votes: 2 2.9%
  • More than 6 weeks

    Votes: 0 0.0%
  • No problems so far

    Votes: 53 75.7%

  • Total voters
    70

infiniteloop

Weaksauce
Joined
Jul 5, 2004
Messages
70
OK, I have seen too many people discussing this, so I think it would be beneficial to everyone if we tried to get a better idea of the numbers on this issue. Please only post if your card is a GTO, I realize that there have been problems with other flavors too.

Also, for you apparently lucky ones that *haven't* had issues, please use the bottom option.

PLEASE NOTE!!

The last option is screwed up, it was meant to be for people who have had the card for 6 weeks already, *and* have still not had any issues. If you've only had it for a few weeks or so, well....you are still in the danger zone. 6+ weeks of ownership would put you in the first wave of purchasers (I think), so having a card that has held up this "long" would be a better reflection of their reliability.
 
Took almost 2 weeks for my card to start artifacting on me. My card had infinion ram too :eek:
 
I only had one problem with artifacting, and it was during a multiplayer game of Railroads! My roomate with a GTO had the problem too, so I definitely think something was going on there... but it's the only problem I've had so I'm marking no problems yet; I don't think it's dying.

Have had the card for about a month BTW.
 
i have two for SLI and had the secondary go pooper last night. YAY! other one seems fine, slightly less than a month here but i work a lot so i havn't been able to test much until last night. i guess i can survive with one until i get the secondary repaired/replaced.
 
I wonder if I should have waited longer than 2 weeks to vote...no problems when I voted, then today I got what can best be described as some "psychadellic" textures in bf2142...but, hasnt done it since, and the clocks are still up, time will tell!
 
7900GTO...... *sighs* :(

hey, infiniteloop, i suspect that we may have seen some other interesting results if the last option, "no problems so far," would have read "no problems after 6 weeks," or something like that. I suspect some of the voters haven't had it too long, but that's probably just me being influenced by my bad experience. :(

But things are looking sunny again as I plan to get an X1900XT 512MB when their prices drop on G80 release. :cool:
 
i won't call my experience a bad one until the second card keels over and puffs out some smoke. evga were very quick about it - they approved my RMA in under an hour!
 
gorilla show said:
hey, infiniteloop, i suspect that we may have seen some other interesting results if the last option, "no problems so far," would have read "no problems after 6 weeks," or something like that.....:

crap, you're right. that's actually what I was intending, and it's what I was thinking when I set up the option. I don't suppose there is a way to change it :(
 
I have a pair. Which option do I choose if they never worked right but it apparently isn't a hardware problem so EVGA won't help me out?
 
dalef said:
I have a pair. Which option do I choose if they never worked right but it apparently isn't a hardware problem so EVGA won't help me out?

That would be the "screwed" column... :(
 
I think I got one of the newer ones. I got mine on the 17th of October and I didn't get one of the "GTX" boxed ones. I'm using the stock cooling and haven't removed it to see what type of RAM it has. I had the "psychedelic" issues with very odd colors in COD 2 and BF2 when I had the ram over 1600. I clocked it back down to 695/1450 and everything has been running great for the last week 1/2. Still looking for another one to go SLI.
 
I've had mine for three weeks, and I'm glad to say there's no issues at all. I got the card running at 700/800 stable too :D
 
Mine's running 720/800 Now, no artifacting detected in ATItool...but...I have a question for those of you who've had failures!

What does a failure look like???

I'm getting some weirdness in BF2142, but so far that is the ONLY game. Cannot get weirdness in Oblivion, HL2-EP1, Middle Earth II, anything. Desktop is fine, temps under load hit 65 degrees MAX ever logged.

The weirdness, for lack of a better term, only occurs in 3 situations, and ALWAYS happens. Will take a screenshot next time it happens, but it can be described as follows:

1 - The dumpsters. Yep. The dumpsters are as if they have no texture, and have a chromy messed up feel to them, showing a smear of what was last on the screen before moving onto the dumpster.

2 - Certain doors in a SINGLE level; same issue as the dumpsters

3 - There is what can only be described as a bloody mess on the Titan ship corridors, right side only. It is like Textures and 3d rendered shapes everywhere, crossing through the walls, and once inside you cant really see anything else aside from the garbage.

Somebody please allow me some mental relief and let me know these kinds of things are not my video card, and just BF2142 being crappy. I should mention these problems are 100% consistant, and seem to affect nothing else...
 
Im wondering what is the difference in the gto and the gtx besides bios and ram speed and timeings. Like what type of ram do they use and so forth. And does the gtx tend to be more reliable.
 
Gee, Nvidia takes the early 7900GTX's that were having trouble with memory failure at advertised clocks, underclocks the memory, renames them GTO's and sells them at a reduced price. People buy the cards, clock them at 7900GTX speeds or faster and they (*GASP*) fail!

How could anyone have predicted this? :rolleyes:
 
JackAshley said:
What does a failure look like???

Here are some screenshots of the kind of failure that made the 7900GT's [in]famous :
http://www.hardforum.com/showpost.php?p=1029959294&postcount=15

Mine had the same simptoms, but much more aggresive (tens or hundreds of spikes (misplaced vertex)).


JackAshley said:
... showing a smear of what was last on the screen before moving onto the dumpster ... let me know these kinds of things are not my video card, and just BF2142 being crappy. I should mention these problems are 100% consistant, and seem to affect nothing else...

I had similar problems when using incorrect brushes mapping for Quake III (well, when trying to learn to map :p ), and that was in the good old times when hardware was reliable (when there was no XFX).

I bet your problems are software related (BF2142 mappers as bad as I was, or a corrupted install).
 
Rabid Badger said:
Gee, Nvidia takes the early 7900GTX's that were having trouble with memory failure at advertised clocks, underclocks the memory, renames them GTO's and sells them at a reduced price. People buy the cards, clock them at 7900GTX speeds or faster and they (*GASP*) fail!

How could anyone have predicted this? :rolleyes:

i'm getting a little tired of people making asumptions like these. how do you know that these were simply failed GTX cards? that doesn't even make sense, since the GPU on the GTO and GTX are identical! secondly, when was nvidia ever having trouble with early GTX cards? only confirmed thing i ever saw was a bit higher than normal failure rates on the 7900 GT's. think you can provide any actual facts here?
 
PutoVicio said:
Here are some screenshots of the kind of failure that made the 7900GT's [in]famous :
http://www.hardforum.com/showpost.php?p=1029959294&postcount=15

Mine had the same simptoms, but much more aggresive (tens or hundreds of spikes (misplaced vertex)).

I bet your problems are software related (BF2142 mappers as bad as I was, or a corrupted install).

erp...it seems like those spikies are what mine has, but in the thousands! Only in a single place in the game though, and not in any other games...I'll pull some screens out of my arse when i get home from work/school, and post...should be amusing :)
 
THRESHIN said:
i'm getting a little tired of people making asumptions like these. how do you know that these were simply failed GTX cards? that doesn't even make sense, since the GPU on the GTO and GTX are identical! secondly, when was nvidia ever having trouble with early GTX cards? only confirmed thing i ever saw was a bit higher than normal failure rates on the 7900 GT's. think you can provide any actual facts here?

First, I didn't say these are cards that have failed, I said they are cards that are likely to fail at 7900GTX speeds that were underclocked and renamed.

Second, it's not the cores that fail, it's the memory.

Check these out:

http://enthusiast.hardocp.com/article.html?art=MTA2OSwsLGhlbnRodXNpYXN0

^^ An entire article on this site that never once specifies the 7900GT.

http://www.hardforum.com/showthread.php?t=1059136

^^ The thread related to that article with no shortage of reports of failing 7900GTX cards.

You were saying?
 
Rabid Badger said:
First, I didn't say these are cards that have failed, I said they are cards that are likely to fail at 7900GTX speeds that were underclocked and renamed.

Second, it's not the cores that fail, it's the memory.

Check these out:

http://enthusiast.hardocp.com/article.html?art=MTA2OSwsLGhlbnRodXNpYXN0

^^ An entire article on this site that never once specifies the 7900GT.

http://www.hardforum.com/showthread.php?t=1059136

^^ The thread related to that article with no shortage of reports of failing 7900GTX cards.

You were saying?

every peice of information you ahve given here is compeletely irrelevant. i'm well aware of the 7900 article on this site. i'm also aware that this article pertains ONLY to factory overclocked cards. to my knowledge, there is no factory overclocked GTO card. we can both agree that the only relevant difference between the GTX and the GTO is memory speed and possibly manufacturer.

so you are stating that it is the memory failing on the GTO cards. care to back that statement up? sure i agree that it very well could be. however, there is nothing here to show me that it is likely the memory. secondly, why are you citing an article that was written before the GTO was even thought of? what good does that do us? same with the thread that sprung up from the article. again, nothing what to do with the GTO. all i'm saying is prove to me that the GTO is a GTX with memory that failed to meet the GTX spec and was underclocked to achieve stability. until then, as far as i'm concerned the GTO could be a card that uses the GTX reference board, same GPU, but different memory/timings than the GTX cards ever used.

now, do you know where to find exact memory information on the GTX (ie. what ram was used in production from start to current) and what is used on the GTO? that information might serve some purpose here...
 
Yep I got issues with O/Cing the card

At stock speeds its fine.

If I got a few mhz faster BOOM crazy artifacting like in the screenshots above.

Weaksauce, .. at least its ok at stock speeds and plays bf2 just fine. This is a "make-due" card for me as I was building a new PC and need something immediately. I got my sights on the new g80 cards in a few months once bugs and prices are ironed out.
 
THRESHIN said:
every peice of information you ahve given here is compeletely irrelevant. i'm well aware of the 7900 article on this site. i'm also aware that this article pertains ONLY to factory overclocked cards. to my knowledge, there is no factory overclocked GTO card. we can both agree that the only relevant difference between the GTX and the GTO is memory speed and possibly manufacturer.

so you are stating that it is the memory failing on the GTO cards. care to back that statement up? sure i agree that it very well could be. however, there is nothing here to show me that it is likely the memory. secondly, why are you citing an article that was written before the GTO was even thought of? what good does that do us? same with the thread that sprung up from the article. again, nothing what to do with the GTO. all i'm saying is prove to me that the GTO is a GTX with memory that failed to meet the GTX spec and was underclocked to achieve stability. until then, as far as i'm concerned the GTO could be a card that uses the GTX reference board, same GPU, but different memory/timings than the GTX cards ever used.

now, do you know where to find exact memory information on the GTX (ie. what ram was used in production from start to current) and what is used on the GTO? that information might serve some purpose here...

Just because your cards havent failed doesn't mean the the issue doesn't exist. My memory failed, so have 20 others out of 60. 1/3 of the GTO(people that voted and that i personally know of) have failed. Not a good track record. Quit defending eVGA and MSI for trying to get rid of Samsung based 7900GTXs. No matter what they do with the memory, undervolt it, underclock it, nothing will fix the fact that the samsung modules ARE deffective. Original run GTXs and later run GTXs are Infineon based and have less chance of failing(if any?) but the middle GTXs are Samsung and they fail. That was the fiasco.
 
Endurancevm said:
Just because your cards havent failed doesn't mean the the issue doesn't exist. My memory failed, so have 20 others out of 60. 1/3 of the GTO(people that voted and that i personally know of) have failed. Not a good track record. Quit defending eVGA and MSI for trying to get rid of Samsung based 7900GTXs. No matter what they do with the memory, undervolt it, underclock it, nothing will fix the fact that the samsung modules ARE deffective. Original run GTXs and later run GTXs are Infineon based and have less chance of failing(if any?) but the middle GTXs are Samsung and they fail. That was the fiasco.

i'm not defending anyone here! i think the failure rates are pathetic- or what little we can preceive of the overall failrate from this thread. very skewed data indeed considering that people with a failed card are more likely to complain that those with a good one. in fact, i'm rather pissed that one of my GTO's died and i had to spend $27 to mail the damn thing back to evga!

my point is this - you are stating that these GTO's are GTX cards that had memory failures and were clocked down, undervolted, etc. to achieve some half-assed stability. now, once again, i'm saying PROVE IT.

simply show me PROOF that the cards have the SAME memory chips. for example, show me a GTO that has memory modules that are also (or were) used on a GTX. those chips will therefore be raited to run at 1600Mhz DDR on both cards regardless of what they are actually running.
 
Back
Top