Reason found: It's one of that 5970's core dead, now I completely disabled that core in cgminer, then the mining rig keeps running as usual
Actually I also experienced once crash in one of my most stable 5870s, I guess there is a limit how many hours you could run them at 99%, should be much less than factory specs
Topic: -18 C degree, mining rig crash ? (Read 1755 times)
Someone else posted here recently who put his rig outside, and his videocards would overheat once the heatpipes froze.
MOAR VOLTAGE NAO 
Temperature in the case will likely stay fairly reasonable with a couple of 5970s. Just turn off some casefans if need be. And if for some reason you must run windows on a dedicated miner, there is always SSDs.
The only real concern Id have (aside from hdds) is heatpipes. Someone else posted here recently who put his rig outside, and his videocards would overheat once the heatpipes froze.
The only real concern Id have (aside from hdds) is heatpipes. Someone else posted here recently who put his rig outside, and his videocards would overheat once the heatpipes froze.
Nice, P4, but do mind that usb+linux miners are - alas! - not the majority.
While there's nothing preventing one from using Windows the same way, all that background housekeeping stuff would kill a usb drive fast.
My criticism still stands, hard drives or no hard drives.
While there's nothing preventing one from using Windows the same way, all that background housekeeping stuff would kill a usb drive fast.
My criticism still stands, hard drives or no hard drives.
Whats a hard drive?
I hate to break this to you man, computers are designed as indoor devices.
Running them at temperatures below condensation point is just asking for trouble.
I don't care what's wrong with the GPU, all that really matters is that you're running your machine far out of spec. Thus undefined behavior should barely be a surprise.
I wouldn't dare to run hard disks in such low temperatures: the grease inside them thickens and you risk mechanical failure.
Running them at temperatures below condensation point is just asking for trouble.
I don't care what's wrong with the GPU, all that really matters is that you're running your machine far out of spec. Thus undefined behavior should barely be a surprise.
I wouldn't dare to run hard disks in such low temperatures: the grease inside them thickens and you risk mechanical failure.
Hmmm.. if its registering 50C, that kinda rules out overheating. Im not intimately familiar with the 5970 cooler design, but I assume it uses a single large cooler, so that may not be it; there is no reason the temp sensor would be lying. Although.. can you monitor VRM temperatures? Perhaps the fan is spinning very slowly because the GPU is so cool, that the basplate covering the VRMs gets too hot? If thats the case, it could be easy to solve, just manually set a higher fan speed.
Other thought, could it be the CPU overheating if that has a heatpipe cooler?
Other thought, could it be the CPU overheating if that has a heatpipe cooler?
CPU running on stock cooler without heatpipe
I think maybe it has something to do with that 5970 card, since it has been working badly after I purchased it(second hand), it always crashed if I raise it above 750Mhz and recently I could only get a stable run at 710Mhz or lower. If there is a way to raise the voltage, maybe I can get a more stable result, but it seems cgminer can not push it over the BIOS limit
Hmmm.. if its registering 50C, that kinda rules out overheating. Im not intimately familiar with the 5970 cooler design, but I assume it uses a single large cooler, so that may not be it; there is no reason the temp sensor would be lying. Although.. can you monitor VRM temperatures? Perhaps the fan is spinning very slowly because the GPU is so cool, that the basplate covering the VRMs gets too hot? If thats the case, it could be easy to solve, just manually set a higher fan speed.
Other thought, could it be the CPU overheating if that has a heatpipe cooler?
Other thought, could it be the CPU overheating if that has a heatpipe cooler?
Ironically, your card is likely overheating. Its because the heatpipes dont work at too low temperature., the liquid in it freezes or no longer evaporates.
If that's the case, I'm relaxed
just put the a box on it will bring the case temperature back... Anyway, the sensor still shows 50 C degree for each GPU when in mining, I doubt that maybe some of the other components in the power supply module could not provide enough voltage due to low temperature
Ironically, your card is likely overheating. Its because the heatpipes dont work at too low temperature., the liquid in it freezes or no longer evaporates.
Is there anyone experienced hardware crash due to very low temperature?
I used to put my machine on balcony and after the air temperature has dropped below 0 degree, I had encounted frequently error in one of the cards (5970). Yesterday temperature hit -18c degree and no matter how much I lower GPU frequency (600 Mhz currently), it will still crash after start cgminer for about 5 minutes
Or, it is just my 5970 runs out of juice?
I used to put my machine on balcony and after the air temperature has dropped below 0 degree, I had encounted frequently error in one of the cards (5970). Yesterday temperature hit -18c degree and no matter how much I lower GPU frequency (600 Mhz currently), it will still crash after start cgminer for about 5 minutes
Or, it is just my 5970 runs out of juice?
Jump to: