2x6970's Crashing Repeatedly with GUIMiner

[Bananington]

I'm trying to mine with two MSI Radeon 6970 Lightning Editions and my computer GSOD's after 5-10 seconds of starting the miner. The crash is consistent.

-The Cards are getting about 800 mhash/s total. I love it.

-I'm using GUIMiner with 2 separate OpenCL miners for my two MSI Radeon 6970 Lightning Editions.

-My flag is -v for both OpenCL miners.

How do I stop this?

[1715175647]

1715175647

[jjiimm_64]

I'm trying to mine with two MSI Radeon 6970 Lightning Editions and my computer GSOD's after 5-10 seconds of starting the miner. The crash is consist.

-The Cards are getting about 800 mhash/s total. I love it.

-I'm using GUIMiner with 2 separate OpenCL miners for my two MSI Radeon 6970 Lightning Editions.

-My flag is -v for both OpenCL miners.

How do I stop this?

lower your clocks until it doesn't crash....  better yet, start them at a low clock and run them for a while.  I do not put my 6970's higher then 900.  if they are new, you should probably burn them in at 800 or 850 for a while.

[the joint]

I'm trying to mine with two MSI Radeon 6970 Lightning Editions and my computer GSOD's after 5-10 seconds of starting the miner. The crash is consist.

-The Cards are getting about 800 mhash/s total. I love it.

-I'm using GUIMiner with 2 separate OpenCL miners for my two MSI Radeon 6970 Lightning Editions.

-My flag is -v for both OpenCL miners.

How do I stop this?

lower your clocks until it doesn't crash....  better yet, start them at a low clock and run them for a while.  I do not put my 6970's higher then 900.  if they are new, you should probably burn them in at 800 or 850 for a while.

I second this.  I was running my 6970 over 900 for a good while until I started getting random crashes.  Not always BSODs, but crashes nonetheless.  I think I ended up sending a power surge through the comp that damaged the cpu and motherboard. Luckily these were replaced for free.

I set the clock back to 850 and am running it there at 335 mhash.

[Bananington]

lower your clocks until it doesn't crash....  better yet, start them at a low clock and run them for a while.  I do not put my 6970's higher then 900.  if they are new, you should probably burn them in at 800 or 850 for a while.

I'll try running this tonight at 900MHz while I sleep and see if it's still running when I wake up. Thank you both. You and the joint.

[Bananington]

Okay, so I'm getting things stable. Now I am having EXTREME temperature issues with 1 GPU. I can lower the Clock voltage, but I would like to know if I can damage my card with low voltages before I do so. How low can I go before I damage the card, if at all?

[the joint]

Okay, so I'm getting things stable. Now I am having EXTREME temperature issues with 1 GPU. I can lower the Clock voltage, but I would like to know if I can damage my card with low voltages before I do so. How low can I go before I damage the card, if at all?

What's your fan setting and temp?

[Bananington]

Okay, so I'm getting things stable. Now I am having EXTREME temperature issues with 1 GPU. I can lower the Clock voltage, but I would like to know if I can damage my card with low voltages before I do so. How low can I go before I damage the card, if at all?

What's your fan setting and temp?

Fan setting is at 100% and temps are going above 100c until it crashes. Once I noticed, I was horrified.

So my question is how low can my Clock voltage go at 875MHz Clock speeds? I currently have it running at 1.0V without a crash yet.

[the joint]

Okay, so I'm getting things stable. Now I am having EXTREME temperature issues with 1 GPU. I can lower the Clock voltage, but I would like to know if I can damage my card with low voltages before I do so. How low can I go before I damage the card, if at all?

What's your fan setting and temp?

Fan setting is at 100% and temps are going above 100c until it crashes. Once I noticed, I was horrified.

So my question is how low can my Clock voltage go at 875MHz Clock speeds? I currently have it running at 1.0V without a crash yet.

Yikes...

Sure you don't have a bad card?

Maybe someone with a little more knowledge can help you.  I haven't messed with the voltage settings much.

[Bananington]

Okay, so I'm getting things stable. Now I am having EXTREME temperature issues with 1 GPU. I can lower the Clock voltage, but I would like to know if I can damage my card with low voltages before I do so. How low can I go before I damage the card, if at all?

What's your fan setting and temp?

Fan setting is at 100% and temps are going above 100c until it crashes. Once I noticed, I was horrified.

So my question is how low can my Clock voltage go at 875MHz Clock speeds? I currently have it running at 1.0V without a crash yet.

Yikes...

Sure you don't have a bad card?

Maybe someone with a little more knowledge can help you.  I haven't messed with the voltage settings much.

I believe it is my case as well as my fans. This is a gaming rig, not a dedicated mining rig.

[P4man]

Lowering voltage will not harm your card. Quite on the contrary, the lower the voltage (and temps and to a lesser extend, clocks) the longer it will live. High temps will kill your card. And it shouldnt be at 100C, or anywhere near. I am assuming you mean dispIO temperature or shader temperature, not VRM temps. Though its still high, VRMs might get that hot and live..

As a quick test, clock everything back to stock. Open the case, en perhaps even point a deskfan at it. What are your temps now? If its still 90C+ you may have to reapply thermal paste.

Another possibility, if you are using GPU-Z and everest and/or afterburner, do not use them simultaneously. I burned a 5850 by running those apps. For some reason it causes the vcore to randomly spike to 1.65v (!). It didnt live long after that.

[Bananington]

Lowering voltage will not harm your card. Quite on the contrary, the lower the voltage (and temps and to a lesser extend, clocks) the longer it will live. High temps will kill your card. And it shouldnt be at 100C, or anywhere near. I am assuming you mean dispIO temperature or shader temperature, not VRM temps. Though its still high, VRMs might get that hot and live..

As a quick test, clock everything back to stock. Open the case, en perhaps even point a deskfan at it. What are your temps now? If its still 90C+ you may have to reapply thermal paste.

Another possibility, if you are using GPU-Z and everest and/or afterburner, do not use them simultaneously. I burned a 5850 by running those apps. For some reason it causes the vcore to randomly spike to 1.65v (!). It didnt live long after that.

I'm only using MSI Afterburner to manage my cards. The cards were both at stock speeds when my was computer crashing. Temps didn't even have a chance to raise in the 3-6 seconds the miner ran.

When I downclock 1 of the cards from stock speeds(the GPU that gets above 100c), like the two above suggested, the miner runs stably(as far as I can tell in less than an hour of constant mining, not enough time though) this gives one of my GPU's time to get above 100c, the other GPU sits at 80C

I've decided to leave one of the GPU's(the one that stays at 80) at stock and I decreased Memory Frequency as low as it'll go(runs stable), and turn down the Aux Voltage(any idea what this controls?), and I decreased the temps in that one by a couple degrees.

The other GPU I have yet to tweak, they both seem to have different sensitivities so I will separately customize their settings.

NOTE: I am using stock thermal paste. I have not done anything but install them into my computer. No modding.

[heinz]

Try cgminer and its overclock and autofan features.

[Bananington]

Try cgminer and its overclock and autofan features.

I have been meaning to get around to trying that. I might do that. But not to fix this issue.

[heinz]

you set your target temp and range of overclock, it finds its sweet spot

[P4man]

When I downclock 1 of the cards from stock speeds(the GPU that gets above 100c), like the two above suggested, the miner runs stably(as far as I can tell in less than an hour of constant mining, not enough time though) this gives one of my GPU's time to get above 100c, the other GPU sits at 80C

You are going to kill your cards at those temps. Find out if something is wrong with the card itself by switching their position (if its an airflow issue) and/or do as I suggested, open the case and point a desk fan at it.

You can tweak voltages and clocks all you want later, but at stock speed and settings, these cards should not run at 100C.

[DeathAndTaxes]

Agreed on the temps and load is unsustainable.  100% fan load 24/7 is going to kill the fan and >100C core temp continually is going to try the GPU.

If you are getting that high of temps @ 100% fan usage either you have the worst case ever built or your card is defective/damaged.

Here is something to try.  Run cards @ stock w/ 100% fan.  Remove side panel on computer case so it can get sufficient cool air.   If core temp stable at <70C?  If so then you need more airflow.  Your case is too restrictive.  Either go to open rig or look into a case with good airflow (think multiple 140m intake fans). 

If you still have crashing and 100C @ stock w/ side panel open ... THE CARD IS DEFECTIVE/DAMAGED.  Period.  It shouldn't be doing that.  If it is under warranty then RMA it before you cook it and they decide to not cover it as user damage.

Just because somebody can do xxx Mhz doesn't mean you can.  A closed system is going

[Bananington]

Calm down. I opened up the case and the 100c card is getting 80c, and the 80c card is getting 70's-60's. It's just my case. One card runs great. But my second card is close to the bottom with minimal air flow.

I've got my mining pretty much stable. I'm tweaking the voltages lower to see what is stable to get cooler temps maxed out.

This rig isn't going to be mining 24/7. It'll be mining overnight or when I'm in class. So it'll be running less than 24 hours per day at those temps and fan speeds. I have a 3 year warranty. Will that cover me if my fans malfunction? I would guess so. I'm not worried about cooking my GPU because anything above 90c is completely unacceptable to me. I try for 70'sc.

[mc_lovin]

Congrats Bananington, you got some cards!!!

[Bananington]

you set your target temp and range of overclock, it finds its sweet spot

That sounds dreamy. I'll try it tonight.

[Bananington]

Congrats Bananington, you got some cards!!!

I also bought a 32" HDTV with full HD 1920x1080 Resolution. My gaming rig is complete.