XFX 7950 Double D Caught on Fire

[PeerMedia]

I have a small LTC rig with 2 XFX Double D 7950's (Black Edition), they run at 81c, slightly overclocked but runs well with an external fan. Great hash rate for a 7950, 730k/card. I came home on Friday to find my rig off. Not sure why (wasn't a power failure), so I went to turn the rig back on and bam, fire started in one of my 7950's. I turned off the power immediately and blew on the GPU until it went out. Scary as hell.

After I let things cool off, I wasn't able to turn on the rig, I was able to restore it when I removed the GPU that caught on fire. The other GPU seems ok, has anyone experienced something like this? I'm petrified to mine again if another fire happened when I'm not home. I have no explanation what happened or how it would catch on fire from turning the power on.

I realize XFX doesn't have a great reputation among miners but I'm stuck on what to possibly do.

Rig specs:
2 XFX Double D, 7950 (Black Edition)
1000W Gold PSU
ASRock Extreme 4
4 GB RAM

Anyone have advice?

[Scyntech]

anything over 75c isn't good. That said, it very well may have been a faulty card. XFX I understand are not great at mining, but they are good gaming cards. I'd suggest dropping your OC a little and play around with the mining program settings and see how high you can get you hash rate up without going over 75c. Also you NEED proper cooling if your going OC and do the normal cleaning...dust and such.

[empoweoqwj]

I have a small LTC rig with 2 XFX Double D 7950's (Black Edition), they run at 81c, slightly overclocked but runs well with an external fan. Great hash rate for a 7950, 730k/card. I came home on Friday to find my rig off. Not sure why (wasn't a power failure), so I went to turn the rig back on and bam, fire started in one of my 7950's. I turned off the power immediately and blew on the GPU until it went out. Scary as hell.

After I let things cool off, I wasn't able to turn on the rig, I was able to restore it when I removed the GPU that caught on fire. Clearly the GPU is fried, the other GPU seems ok, has anyone experienced something like this? Even if I can RMA this GPU, I'm petrified to mine again if another fire happened when I'm not home. I have no explanation what happened or how it would catch on fire from turning the power on.

I realize XFX doesn't have a great reputation among miners but I'm stuck on what to possibly do.

Rig specs:
2 XFX Double D, 7950 (Black Edition)
1 1000W Rosewill Gold PSU
ASRock Z77 Extreme4
4 GB G.Skill DDR3 RAM

Anyone have advice?

Its always a possibility when running electrical devices "hot". Strange it happened just when you turned the device on though. No idea why that would happen .... probably result of previous mining session. Did you ever measure card temperatures before the "big bang"?

[PeerMedia]

I cannot for the life of me get the temperature lower than 80c even with an industrial fan (identical temperatures without or with the fan). I always assumed their temperature was calculated by fan RPMs and not actually measured at the source.

I've backed off on intensity and OC a bit on the other GPU but temperature is still 79c, not much cooler.

[Scyntech]

I cannot for the life of me get the temperature lower than 80c even with an industrial fan (identical temperatures without or with the fan). I always assumed their temperature was calculated by fan RPMs and not actually measured at the source.

I've backed off on intensity and OC a bit on the 1 GPU that hasn't fried but temperature is still 79c, not much cooler.

It's possible the crappy thermal paste (most companies) they use needs to be replaced OR the heatsink fins are dirty. How hot does it run at idle?

[PeerMedia]

I cannot for the life of me get the temperature lower than 80c even with an industrial fan (identical temperatures without or with the fan). I always assumed their temperature was calculated by fan RPMs and not actually measured at the source.

I've backed off on intensity and OC a bit on the 1 GPU that hasn't fried but temperature is still 79c, not much cooler.

It's possible the crappy thermal paste (most companies) they use needs to be replaced OR the heatsink fins are dirty. How hot does it run at idle?

I don't have many ways to check the temps other than the first reading when starting cgminer, in this case its around 40c. The GPU's were just 5 weeks old, very clean room, no dust, no pets. I also wanted to add that my rig is in a milk crate and very well spaced out with risers.

[Scyntech]

http://event.msi.com/vga/afterburner/download.htm MSI Afterburner. That will tell you the temps. If those are only 5 weeks old. RMA them, you shouldn't be having these issues.

[PeerMedia]

http://event.msi.com/vga/afterburner/download.htm MSI Afterburner. That will tell you the temps. If those are only 5 weeks old. RMA them, you shouldn't be having these issues.

I'm on Xubuntu so MSI won't work (I think?)

[empoweoqwj]

http://event.msi.com/vga/afterburner/download.htm MSI Afterburner. That will tell you the temps. If those are only 5 weeks old. RMA them, you shouldn't be having these issues.

Yep, you defininitely need to measure the temp whilst running. Then you will know at what stage things start to go wrong.

[Scyntech]

http://event.msi.com/vga/afterburner/download.htm MSI Afterburner. That will tell you the temps. If those are only 5 weeks old. RMA them, you shouldn't be having these issues.

I'm on Xubuntu so MSI won't work (I think?)

I've never used Ubuntu, quick Google search here https://www.google.com/search?q=Xubuntu&oq=Xubuntu&aqs=chrome..69i57&sourceid=chrome&espv=210&es_sm=122&ie=UTF-8#q=ubuntu+gpu+temp+monitor&safe=off

I don't think MSI has a temp monitor for Ubuntu

[empoweoqwj]

http://event.msi.com/vga/afterburner/download.htm MSI Afterburner. That will tell you the temps. If those are only 5 weeks old. RMA them, you shouldn't be having these issues.

I'm on Xubuntu so MSI won't work (I think?)

I've never used Ubuntu, quick Google search here https://www.google.com/search?q=Xubuntu&oq=Xubuntu&aqs=chrome..69i57&sourceid=chrome&espv=210&es_sm=122&ie=UTF-8#q=ubuntu+gpu+temp+monitor&safe=off

I don't think MSI has a temp monitor for Ubuntu

There's a serious issue. How can you run a rig without monitoring temp? Scary stuff....

[Scyntech]

http://event.msi.com/vga/afterburner/download.htm MSI Afterburner. That will tell you the temps. If those are only 5 weeks old. RMA them, you shouldn't be having these issues.

I'm on Xubuntu so MSI won't work (I think?)

I've never used Ubuntu, quick Google search here https://www.google.com/search?q=Xubuntu&oq=Xubuntu&aqs=chrome..69i57&sourceid=chrome&espv=210&es_sm=122&ie=UTF-8#q=ubuntu+gpu+temp+monitor&safe=off

I don't think MSI has a temp monitor for Ubuntu

There's a serious issue. How can you run a rig without monitoring temp? Scary stuff....

Agreed, I've seen guys lose $1500+ gaming rigs because of either hardware failures or improper overclocking. I've killed a few Opteron(s) in my day 

[PeerMedia]

CGMiner has temperature monitoring through aticonfig and I setup cgminer temps to 80 (target), 83 (overheat), 85 (shutoff). It never exceeded 82c while running and I regularly monitor temps.

[empoweoqwj]

http://event.msi.com/vga/afterburner/download.htm MSI Afterburner. That will tell you the temps. If those are only 5 weeks old. RMA them, you shouldn't be having these issues.

I'm on Xubuntu so MSI won't work (I think?)

I've never used Ubuntu, quick Google search here https://www.google.com/search?q=Xubuntu&oq=Xubuntu&aqs=chrome..69i57&sourceid=chrome&espv=210&es_sm=122&ie=UTF-8#q=ubuntu+gpu+temp+monitor&safe=off

I don't think MSI has a temp monitor for Ubuntu

There's a serious issue. How can you run a rig without monitoring temp? Scary stuff....

Agreed, I've seen guys lose $1500+ gaming rigs because of either hardware failures or improper overclocking. I've killed a few Opteron(s) in my day 

I'm still wondering why it caught fire when the OP turned it back on though. I suppose something got fried when he was out, and the system luckily cut out instead of catching on fire then. I think the OP got off very lucky. He could have come back to a shell of a house.

[PeerMedia]

anything over 75c isn't good. That said, it very well may have been a faulty card. XFX I understand are not great at mining, but they are good gaming cards. I'd suggest dropping your OC a little and play around with the mining program settings and see how high you can get you hash rate up without going over 75c. Also you NEED proper cooling if your going OC and do the normal cleaning...dust and such.

I've cut all OC out and running at stock settings, its running at 79c with the external fan, saved just 2 degrees by not OC'ing and lost 100kh/s.

[Scyntech]

CGMiner has temperature monitoring through aticonfig and I setup cgminer temps to 80 (target), 83 (overheat), 85 (shutoff). It never exceeded 82c while running and I regularly monitor temps.

Do some research on your cards. See if those have general heat/overheating issues.

 My 7850 is OC'ed from 860Mh/z to 1100Mh/z core and 1200Mh/z to 1350 Mh/z Mem. At 99% load I never reach 76c with a 2% increase in voltage. It's also fan cooled, I have two fans on the back pulling air out of the case. Two on the side and two in front pushing air in.

I know XFX and a few off brands run hot, but that doesn't me ALL XFX cards will have heat issues.

[Scyntech]

Kind of what I thought

http://hardforum.com/showthread.php?t=1700706

there are a number of threads of people asking about high temps with the same card you have. Reviews show it should be running at much lower temps. Thinking you got a couple of cards from a bad batch. It does happen.

[Gator-hex]

The other GPU seems ok, has anyone experienced something like this? I'm petrified to mine again if another fire happened when I'm not home.

Fires are caused by dust balls building up in the PSU, CPU and GPU heatsink and fans. Buy a can of compressed air and clean them out once a month. Put a smoke alarm near your equipment and never place equipment where fire can spread, near curtains etc.

[empoweoqwj]

The other GPU seems ok, has anyone experienced something like this? I'm petrified to mine again if another fire happened when I'm not home.

Fires are caused by dust balls building up in the PSU, CPU and GPU heatsink and fans. Buy a can of compressed air and clean them out once a month. Put a smoke alarm near your equipment and never place equipment where fire can spread, near curtains etc.

Great advice. Is there any material you can "surround" your rig with to prevent fire spreading should it start? Also think about what your rig is standing on e.g. not a carpet that is going to catch fire.

[Gator-hex]

The other GPU seems ok, has anyone experienced something like this? I'm petrified to mine again if another fire happened when I'm not home.

Fires are caused by dust balls building up in the PSU, CPU and GPU heatsink and fans. Buy a can of compressed air and clean them out once a month. Put a smoke alarm near your equipment and never place equipment where fire can spread, near curtains etc.

Great advice. Is there any material you can "surround" your rig with to prevent fire spreading should it start? Also think about what your rig is standing on e.g. not a carpet that is going to catch fire.

Cooking tin foil is a good heat/fire retardant shiney side facing the fire source. It's also a good conductor of electrcity though so be careful not to put it in contact with anything electrical. You can also get fire retardant paints but it's not as cheap as tin foil.