Yep, I have the exact model Sapphire 7950, got it as a bundle deal with a Kingwin 750 W PSU from TigerDirect in December'13, and it's dead. Saw spark near the PCI-E power ports. It won't POST. Emailed and got RMA, it's delivered back for RMA and now waiting for them to confirm.
I also have an older model (11196-16-20G) and that one's been working perfectly since early last year!
Something's definitely wrong with the 11196-19-20G model.
[Edited for clarity and spelling]
Topic: All Sapphire 7950s (11196-19-20G) are not suitable for mining? - page 4. (Read 16872 times)
Yeah sparks i don't put in doubt.
One day, I saw very good the little sparks in one of my dead gpus.
When i said this issue to Sapphire forum, a mod insinuated me that i must have a quality hardware above all the psu.
And I said her that the gpu showed me sparks because i had a bad hardware such as a mobo of 200€ and a Platinum PSU (Fractal Design Newton R3 1000w)
and that if I had a Diamond PSU with 110% eficiency 
i'm sure that i wouldn't see any spark. Sapphire, Sapphire... 
One day, I saw very good the little sparks in one of my dead gpus.
When i said this issue to Sapphire forum, a mod insinuated me that i must have a quality hardware above all the psu.
And I said her that the gpu showed me sparks because i had a bad hardware such as a mobo of 200€ and a Platinum PSU (Fractal Design Newton R3 1000w)
and that if I had a Diamond PSU with 110% eficiency i'm sure that i wouldn't see any spark. Sapphire, Sapphire... Yup, I'm sure many have had problems with their sapphire 7950's of the most recent batch. Something tells me they were created in haste to sell to miners without doing proper quality assurance. The problem is most people don't/won't see this thread but the issue is really bad. Hopefully sapphire issues us new better cards instead of the crappy problematic cards we currently have.
Yeah sparks i don't put in doubt.
One day, I saw very good the little sparks in one of my dead gpus.
When i said this issue to Sapphire forum, a mod insinuated me that i must have a quality hardware above all the psu.
And I said her that the gpu showed me sparks because i had a bad hardware such as a mobo of 200€ and a Platinum PSU (Fractal Design Newton R3 1000w) and that if I had a Diamond PSU with 110% eficiency i'm sure that i wouldn't see any spark. Sapphire, Sapphire...
One day, I saw very good the little sparks in one of my dead gpus.
When i said this issue to Sapphire forum, a mod insinuated me that i must have a quality hardware above all the psu.
And I said her that the gpu showed me sparks because i had a bad hardware such as a mobo of 200€ and a Platinum PSU (Fractal Design Newton R3 1000w)
and that if I had a Diamond PSU with 110% eficiency i'm sure that i wouldn't see any spark. Sapphire, Sapphire... sparks?? really?
OMG...
one would expect Sapphire to recall all affected cards.. but no statement on their website or anything..
OMG...
one would expect Sapphire to recall all affected cards.. but no statement on their website or anything..
Yup sparked when I hit the power on it, I feel like it may be that the card had the voltage turned up on it (1.256) but I never overvolted it at all and can't change it unless i flash the bios, I tried to turn it down but it wouldn't change values. I'm worried about the rest of the cards but I am getting a new power supply and it is a platinum corsair so it shouldn't be a problem. If it sparks or dies i know it would definitely be the defective cards and I would have to do another RMA.
sparks?? really?
OMG...
one would expect Sapphire to recall all affected cards.. but no statement on their website or anything..
OMG...
one would expect Sapphire to recall all affected cards.. but no statement on their website or anything..
I bought six of these cards, two of them have already gone. What I noticed is that they sparked near the PCI-Power connector after I found my system randomly shut down. I think some of these are definitely defective, these were the batch from early January. I feel like it could be the power supply I used so I am trying with different power supplies and seeing if anything happens.
In two weeks my card become dead since I have first started mining. I used it every day except half an hour break (2 times). Most of people says that they burnt their cards. I think this cards are problematic and not suitable for mining?
Did you have the same problems?
My one is ok... i just don't oc it to much Did you have the same problems?
In two weeks my card become dead since I have first started mining. I used it every day except half an hour break (2 times). Most of people says that they burnt their cards. I think this cards are problematic and not suitable for mining?
Did you have the same problems?
Did you have the same problems?
7950 is the worst card ever created.
(* If I had a satoshi for everytime I've posted that comment)
installed the suspected card in my Win 8 and it immediately booted a low res desktop. Nice! right clicked the desktop to set bigger res and immediately black screen and monitor complaining about not getting signal, same symptoms as on Linux so i guess i can safely conclude i fried that card. Did not do any deliberate overclocking or undervolting yet but maybe it was the 74 degrees / I=17 that did it. Anyway the cards were not making any ROI with BTC down (and therefore the alts as well), high power costs over here and running below targets.
My failures gpus turned off my computer and then when i tried to power on again the pc didn't turned on (but removed the conflictives gpus and pc turned on without problems). It was like if a short circuit happened with the cards, perhaps any capacitor but i'm not sure. Curiously i have other cards (same model) working always over 75º temps 24/7 during more than 2 months without problems so i can said than temps & voltage wasn't the issue.
installed the suspected card in my Win 8 and it immediately booted a low res desktop. Nice! right clicked the desktop to set bigger res and immediately black screen and monitor complaining about not getting signal, same symptoms as on Linux so i guess i can safely conclude i fried that card. Did not do any deliberate overclocking or undervolting yet but maybe it was the 74 degrees / I=17 that did it. Anyway the cards were not making any ROI with BTC down (and therefore the alts as well), high power costs over here and running below targets.
Still not a definitive conclusion but i think i burned one two:
Sapphire A1351000 Dual X 3GB with boost (but not runniing boost, or only short time, did not hash any faster with it) out of two deliverd recently (one on the 8th and one on the 11th or so of February) one of them is probably the cause for my system crash last saturday. I did smell something occasionally in the beginning but not lately. Cards were both plugged directly into the pci-e slots. First week i did not run 24x7 as i still had some linux config work to do to get it all working. After that it ran 24x7 with auto-gpu and the highest thread concurrency (around 20k) i could set and ran them in sgminer with I=17 at around 550kH/s. Idle power was 65 watts and mining at 580. Recently i saw that climbing to near 600. The top card had 3000 RPM listed and the bottom one half (could be software as well? ) and they showed 74 degrees Celsius in the software. Top one seems to have died.
At the moment the screen goes black after it tries to launch the Linux graphical desktop, if both are powered i wont get that far. Bottom card on only succesfully boots. Had some trouble with the HD as well until i plugged it in with another sata cable and sata port on my mobo. So maybe i fried the mobo and/or PSU as well. Gonna plug the 7950 into my Win desktop to see what will happen then.
Sapphire A1351000 Dual X 3GB with boost (but not runniing boost, or only short time, did not hash any faster with it) out of two deliverd recently (one on the 8th and one on the 11th or so of February) one of them is probably the cause for my system crash last saturday. I did smell something occasionally in the beginning but not lately. Cards were both plugged directly into the pci-e slots. First week i did not run 24x7 as i still had some linux config work to do to get it all working. After that it ran 24x7 with auto-gpu and the highest thread concurrency (around 20k) i could set and ran them in sgminer with I=17 at around 550kH/s. Idle power was 65 watts and mining at 580. Recently i saw that climbing to near 600. The top card had 3000 RPM listed and the bottom one half (could be software as well? ) and they showed 74 degrees Celsius in the software. Top one seems to have died.
At the moment the screen goes black after it tries to launch the Linux graphical desktop, if both are powered i wont get that far. Bottom card on only succesfully boots. Had some trouble with the HD as well until i plugged it in with another sata cable and sata port on my mobo. So maybe i fried the mobo and/or PSU as well. Gonna plug the 7950 into my Win desktop to see what will happen then.
There seem to have been a broken batch of these shipped during the start of the year, a friend bought a bunch of them and most of them died within weeks. Things like these happens, we just notice them more than "normal" consumers since we buy cards in bulk ;P
wish i had found this thread earlier..
i started out with 3 XFX cards but then read somewhere that sapphire cards are supposed to be the 'quality' brand so i bought an additional 6 sapphire's 7950 dual-x cards
already 5 of those 6 cards are fried in no less than 3 weeks
sku: 11196-19-20g
pn 299-1e249-100sa
all serial numbers between: A135100029900 - A135100030500
first i suspected the risers... but then it also happened to those plugged in directly to mobo..
so then i suspected the motherboard (asrock H81 BTC) but then it also happened with a gigabyte GA-Z87X-OC ..
i also have 3 XFX cards and although they get really hot they work flawless from day 1 in any of those mobo's so conclusion is that the sapphire videocards are flawed..
i started out with 3 XFX cards but then read somewhere that sapphire cards are supposed to be the 'quality' brand so i bought an additional 6 sapphire's 7950 dual-x cards
already 5 of those 6 cards are fried in no less than 3 weeks
sku: 11196-19-20g
pn 299-1e249-100sa
all serial numbers between: A135100029900 - A135100030500
first i suspected the risers... but then it also happened to those plugged in directly to mobo..
so then i suspected the motherboard (asrock H81 BTC) but then it also happened with a gigabyte GA-Z87X-OC ..
i also have 3 XFX cards and although they get really hot they work flawless from day 1 in any of those mobo's so conclusion is that the sapphire videocards are flawed..
February 12, 2014, 09:11:38 PM
What is best ram Hynix or Elpida? because all my dead gpu's had Elpida's ram.
Something weird had my dead gpus because they only cost me 189€ each one and in the online shop (the biggest French online shop) put FLASH OFFERTS Sapphire HD 7950!!
Something weird had my dead gpus because they only cost me 189€ each one and in the online shop (the biggest French online shop) put FLASH OFFERTS Sapphire HD 7950!!
February 12, 2014, 09:45:30 AM
Me too, 5 dual-x 7950 11196-19-20G dead. All of these gpu's had SN: A1350000XXXXX.
And i can confirm that Sapphire 7950 11196-19-20G are ok for mining but Sapphire in the last months has made shits gpu's.
Post here if you have the SN: of your dead gpus i bid that all has a SN like A1350000XXXXX.
I have other SN like A1351000XXXXX and A1345000XXXXX works since 8 months ago and no problems.
And i can confirm that Sapphire 7950 11196-19-20G are ok for mining but Sapphire in the last months has made shits gpu's.
Post here if you have the SN: of your dead gpus i bid that all has a SN like A1350000XXXXX.
I have other SN like A1351000XXXXX and A1345000XXXXX works since 8 months ago and no problems.
Do they are dead almost the same time?
No, and also they continued dead in other computer and with other hardware componentes, so I discard a problem with PSU's or pci-e x1 to x16 risers.
The overclock was minimum, 1050 gpu clock and 1450 memory with 1,15v and temperatures were always good under 70º C.
I think that Sapphire sold refried refurbished cards for mitigate the non stock hd 7950 in Christmas 2013.
My dead gpu's was bought 20th december 2013.
Sapphire has a good rep, this is is weird indeed.
February 12, 2014, 06:45:30 AM
Me too, 5 dual-x 7950 11196-19-20G dead. All of these gpu's had SN: A1350000XXXXX.
And i can confirm that Sapphire 7950 11196-19-20G are ok for mining but Sapphire in the last months has made shits gpu's.
Post here if you have the SN: of your dead gpus i bid that all has a SN like A1350000XXXXX.
I have other SN like A1351000XXXXX and A1345000XXXXX works since 8 months ago and no problems.
And i can confirm that Sapphire 7950 11196-19-20G are ok for mining but Sapphire in the last months has made shits gpu's.
Post here if you have the SN: of your dead gpus i bid that all has a SN like A1350000XXXXX.
I have other SN like A1351000XXXXX and A1345000XXXXX works since 8 months ago and no problems.
Do they are dead almost the same time?
No, and also they continued dead in other computer and with other hardware componentes, so I discard a problem with PSU's or pci-e x1 to x16 risers.
The overclock was minimum, 1050 gpu clock and 1450 memory with 1,15v and temperatures were always good under 70º C.
I think that Sapphire sold refried refurbished cards for mitigate the non stock hd 7950 in Christmas 2013.
My dead gpu's was bought 20th december 2013.
Sounds really weird, maybe their quality control is out due to holidays..
February 11, 2014, 09:41:52 PM
Me too, 5 dual-x 7950 11196-19-20G dead. All of these gpu's had SN: A1350000XXXXX.
And i can confirm that Sapphire 7950 11196-19-20G are ok for mining but Sapphire in the last months has made shits gpu's.
Post here if you have the SN: of your dead gpus i bid that all has a SN like A1350000XXXXX.
I have other SN like A1351000XXXXX and A1345000XXXXX works since 8 months ago and no problems.
And i can confirm that Sapphire 7950 11196-19-20G are ok for mining but Sapphire in the last months has made shits gpu's.
Post here if you have the SN: of your dead gpus i bid that all has a SN like A1350000XXXXX.
I have other SN like A1351000XXXXX and A1345000XXXXX works since 8 months ago and no problems.
Do they are dead almost the same time?
No, and also they continued dead in other computer and with other hardware componentes, so I discard a problem with PSU's or pci-e x1 to x16 risers.
The overclock was minimum, 1050 gpu clock and 1450 memory with 1,15v and temperatures were always good under 70º C.
I think that Sapphire sold refried refurbished cards for mitigate the non stock hd 7950 in Christmas 2013.
My dead gpu's was bought 20th december 2013.
February 11, 2014, 08:17:51 PM
Me too, 5 dual-x 7950 11196-19-20G dead. All of these gpu's had SN: A1350000XXXXX.
And i can confirm that Sapphire 7950 11196-19-20G are ok for mining but Sapphire in the last months has made shits gpu's.
Post here if you have the SN: of your dead gpus i bid that all has a SN like A1350000XXXXX.
I have other SN like A1351000XXXXX and A1345000XXXXX works since 8 months ago and no problems.
And i can confirm that Sapphire 7950 11196-19-20G are ok for mining but Sapphire in the last months has made shits gpu's.
Post here if you have the SN: of your dead gpus i bid that all has a SN like A1350000XXXXX.
I have other SN like A1351000XXXXX and A1345000XXXXX works since 8 months ago and no problems.
Do they are dead almost the same time?
February 11, 2014, 05:02:39 PM
Always use the Sapphire OEM drivers, never the generic AMD ones, Sapphire often do their own auto power management and stuff.
Don't overclock outside what AMD Overdrive recommends as max or you'll burn your card, they recommend it for a reason.
Even if your GPU says 80c your VRMs could be 150+ and trying to melt themselves off the board. Jet a thermometer gun.
Don't overclock outside what AMD Overdrive recommends as max or you'll burn your card, they recommend it for a reason.
Even if your GPU says 80c your VRMs could be 150+ and trying to melt themselves off the board. Jet a thermometer gun.
February 11, 2014, 04:42:48 PM
Me too, 5 dual-x 7950 11196-19-20G dead. All of these gpu's had SN: A1350000XXXXX and A1351000XXXXX.
And i can confirm that Sapphire 7950 11196-19-20G are ok for mining but Sapphire in the last months has made shits gpu's.
Post here if you have the SN: of your dead gpus i bid that all has a SN like A1350000XXXXX.
I have other SN like A1404000XXXXX and A1345000XXXXX working since 8 months ago and no problems.
And i can confirm that Sapphire 7950 11196-19-20G are ok for mining but Sapphire in the last months has made shits gpu's.
Post here if you have the SN: of your dead gpus i bid that all has a SN like A1350000XXXXX.
I have other SN like A1404000XXXXX and A1345000XXXXX working since 8 months ago and no problems.
Jump to: