GPU usage problem - page 2. | Bitcointalksearch.org

DarkMiningBTC

full member

Activity: 156

Merit: 100

Quote from: woutersteven on December 24, 2013, 11:29:22 PM

I had this problem before with my R9 280x Vapor-X cards. In my case it was a difference in the clock speed between the two cards. I'm now using the following settings for CGMiner and that seems to work well:

--gpu-platform 1 -d 1 -w 256 -v 1 -I 13 -g 2 -l 1 -T --lookup-gap 2 --gpu-powertune -20,-20 --auto-fan --temp-target 65,65 --temp-overheat 85,85 --temp-cutoff 90,90 --gpu-memclock 1500,1500 --gpu-engine 1080,1080 --expiry 1 --scan-time 1 --queue 0 --no-submit-stale --thread-concurrency 11200

I'm getting 750-760 khs on both cards with these settings.

ok tried out the settings, didn't fix anything Sad

how do you set the clock speed for BOTH gpus?

DarkMiningBTC

full member

Activity: 156

Merit: 100

well nvm about -T and gpu platform but still doesn't fix it :/

DarkMiningBTC

full member

Activity: 156

Merit: 100

Quote from: woutersteven on December 24, 2013, 11:29:22 PM

I had this problem before with my R9 280x Vapor-X cards. In my case it was a difference in the clock speed between the two cards. I'm now using the following settings for CGMiner and that seems to work well:

--gpu-platform 1 -d 1 -w 256 -v 1 -I 13 -g 2 -l 1 -T --lookup-gap 2 --gpu-powertune -20,-20 --auto-fan --temp-target 65,65 --temp-overheat 85,85 --temp-cutoff 90,90 --gpu-memclock 1500,1500 --gpu-engine 1080,1080 --expiry 1 --scan-time 1 --queue 0 --no-submit-stale --thread-concurrency 11200

I'm getting 750-760 khs on both cards with these settings.

looks like gpu platform 1 is the problem and -T also this is significantly lowering my hash.

DarkMiningBTC

full member

Activity: 156

Merit: 100

Quote from: woutersteven on December 24, 2013, 11:29:22 PM

I had this problem before with my R9 280x Vapor-X cards. In my case it was a difference in the clock speed between the two cards. I'm now using the following settings for CGMiner and that seems to work well:

--gpu-platform 1 -d 1 -w 256 -v 1 -I 13 -g 2 -l 1 -T --lookup-gap 2 --gpu-powertune -20,-20 --auto-fan --temp-target 65,65 --temp-overheat 85,85 --temp-cutoff 90,90 --gpu-memclock 1500,1500 --gpu-engine 1080,1080 --expiry 1 --scan-time 1 --queue 0 --no-submit-stale --thread-concurrency 11200

I'm getting 750-760 khs on both cards with these settings.

CGMiner giving me an error with those settings :/ also arn't you just enabling 1 card like that?

xocel

full member

Activity: 121

Merit: 100

Interesting, almost identical to what im using, but not quite. Will go play around for a bit, see if i can sort it out.

Edit: Never mind, its not doing it atm, must have just cooled down enough, will run some tests tomorrow afternoon.

woutersteven

member

Activity: 89

Merit: 100

I had this problem before with my R9 280x Vapor-X cards. In my case it was a difference in the clock speed between the two cards. I'm now using the following settings for CGMiner and that seems to work well:

--gpu-platform 1 -d 1 -w 256 -v 1 -I 13 -g 2 -l 1 -T --lookup-gap 2 --gpu-powertune -20,-20 --auto-fan --temp-target 65,65 --temp-overheat 85,85 --temp-cutoff 90,90 --gpu-memclock 1500,1500 --gpu-engine 1080,1080 --expiry 1 --scan-time 1 --queue 0 --no-submit-stale --thread-concurrency 11200

I'm getting 750-760 khs on both cards with these settings.

DarkMiningBTC

full member

Activity: 156

Merit: 100

god I can't believe this :/ still trying to figuring out what to do

xocel

full member

Activity: 121

Merit: 100

I have been experiencing this issue with my 280x vapour x's also. The earlier ones dont seem to do it, the newer batches though... Huh

I first noticed they did it around 80C but after using them for a week the temp which initiates the throttling seems to have lowered. first to 78, then 75 now it seems to be around 73C.

I currently have my target temp set to 70 and increased my cooling. Only the card in the top slot does it now, and only during the day. I know these cards can run hotter then this so its very frustrating. I have 2 identical cards from an earlier batch that happily run at 80C with no throttling.

This occurs in both windows and linux, and with different version drivers. I can only guess there's something on the cards causing it to happen.

Edit: Oh, and i have 4 of the vapour x's from the new batch, and they all do it. None of the earlier ones do Undecided

DarkMiningBTC

full member

Activity: 156

Merit: 100

I have 2x 280x's and for the past few days this problem has been seriously pissing me off. Both cards are running 66C (they are xfx DD 280x's) and I have the perfect settings for them running 760KH/s. Though GPU 0's usage goes from 100% down to 64 and back up fluctuating. This causes the hash to be dramastically less. YES I've tried everything, its very cool and has dummy plugs , I even plugged monitors to them and same thing. Please tell me how to fix this Sad

Topic: GPU usage problem - page 2. (Read 1362 times)