Bitcoin Forum

Hi

I just bought an ATI 7970 and ready to mine :)

Right now I got these values:

     Activity: 99%
     GPU clock: 1100%
     Power: 0%
     Temperature: 70%
     Fan Speed: 92%

are these values ok? where can I get more info about temperature and OC?


Thanks! Happy Saturday
B.

Edit: Temperature: 70º C :)

70 FAHRENHEIT is super cool, your PC must be in a fridge.  :)
70 CELSIUS is a good temp, especially with an OC.
70% ??? who knows

Get GPU-Z to read your GPU's temperature sensors.

Its not recommended to mine 24/7 with fan speed at 90%+. Try to keep below 80%. The GPU fan may fail much sooner if you don't.

CGMINER is good for setting fans speeds, clock speeds, etc. It is command line app, you'll need to read up on it.

Yeah I would be happy if my cards would stay around 70'C


Usually closer to 80'C I would start worrying if you start hitting 90'C and above

What do you recommend to not have your fans 24/7 at 90%? what do you do to cool it down?

keeps fans @ 85%...

remember there are multiple GPU sensors...  if thermal paste or pads aren't applied correctly, you could be seeing a 65oC for the card, but gpu sensor 3 might be at 95oC.   i consider it a 'good job' if they're all within 10oC of each other

what's safe for VRM temps varies quite a bit, depending on the manufacturer...  but it *should* be good up to at least 105-110oC.    some can run stable up to 120-130oC.   i've got one card that has been running nonstop (outside of a couple brownouts and one blackout of a few hrs) w/ a VRM temp ranging from 95-105oC.. never crashed once... 5870 @ 1000/315.   gpu temp ranges from 55-65, sensor 3 up to about 70oC.

oh, I run all my fans at 100%.   from my experience, it wears them out quicker w/ variable speeds (i.e. slowing down as the card cools, going faster as it gets hot, etc).    i've been running ~15 cards 24/7 for the last 7 or 8 months and have had to replace 4 fans (3 from reference 5870/5970, and 1 from a gigabyte with that crappy dual fan setup... i just ziptied a 120mm on that)....

and one PSU failure =p

92% fan(s) I'd say it's too high. Try keep your GPU fans under 80%. They'll last much longer. Anyway at 85% seems to be the point of diminishing returns.
Maybe you need to install a front fan to bring fresh air into your case toward your GPU air intake.
70 C is very decent for 1100Mhz. I'm not bothered until 80C from a reliability point of view. But it good to keep temper9ature low for at least two reasons: 1. current consumption is significantly lower; 2, overclockability is better (if you can keep your 7970 at 1125MHz at 80-85C, then probably you can reach 1150 under 70C).

I have a low end GPU from ATI in my desktop machine. Suddenly I was having problems playing quakelive with crashes as soon as I started to play. I ignored this because I assumed it was something to do with the game and gave up caring. A few months go by and I start to get stuttering with video playback on things like Netflix then VLC. So I start to investigate my GPU and find out my GPU was running at over 100C on idle! I opened the case up and cleared dust and it runs now at 50C. I am amazed it didn't brick the GPU because it was running hot for MONTHS.

I've been mining for 4 hours. I overclocked it and put the fans at 85% (manual mode) and got these:

    Activity: 99%
     GPU clock: 1150%
     Power: 0%
     Temperature: 72º C
     Fan Speed: 85%

For what I read, it seems pretty good if I don't reach 85º-90º C, right?

Also I got a new case, so I have more room for fans. 7 total. (not counting power supply and CPU fans)

3 fans installed on the graphic card (came with them)
2 fans at front pulling in air
1 fan at the back pulling out
1 fan on top of the case, pulling out

i have a 6870 and mine runs at around 89 celcius when mining.  fans are at about 95%+

^ you should take that card apart and clean it and apply new thermal paste.


I run 4x cards per board all very close to each other and try to keep things under 80'C

5 more hours mining (9 hours total no-stopping)

    Activity: 99%
     GPU clock: 1150%
     Power: 0%
     Temperature: 69º C (I assume because it's already night and the outside temperature has dropped significantly)
     Fan Speed: 85%

It seems that works pretty well  ;D ;D ;D

let's see how it behaves in its first 24h of life  :D

Edit: Why is "power" 0%? What is that?

Power 0% I believe it's the "powertune" which defines the power envelope of 6900 or newer cards. It throttles down the card if power disipation exceeds TDP, which I believe it's around 250W for 7970. For 6950 with unlocked shaders it was mandatory to set at +20% (the maximum) in order to not throttle. On 7950 and 7970 I did not noticed it's necessary, but just to be sure you can set it at 20%.
It's easy to discover if it's necessary: you look at the hash rate. It's necesary to set if you don't have the  expected hashrate. You should obtain more than 680MH/s at 1150MHz (cgminer 2.5.0, intensity 11, default settings).

Disassembling the card should be the last step, because it's difficult and risky. Before that you should clean card with compressed air, if it looks very dusty. And before that you should have at least a front intake fan in your case  :)

I've had several cards fail at 70-80c. I dropped the temps down to 60-65c and none of them have had issues. They can run for several months even at 80c, especially newer cards. They will eventually go bad though.

My temps are around 60°C to 71°C with fan speed at 80%. So I guess the card and fan would be ok then running 24/7

ATI Sapphire Radeon 7950 OC Edition.

O yeah fan setup as follows:

8 x 24v Fans directing airflow to the card
1 x back of PC vacuum outwards
2 x 240v Fans vacuum from the top
2 x Fans stock standard on the card.

This all with a current heatwave in my country. Max temp today should be 32°c. Yesterday we had a top of 31°c and the card reached that 71° I posted earlier.

Really? It's difficult? I never realized. I have taken apart and cleaned 20+ cards.. I never once had a hard time. Make sure you have the proper tools.

I have 3 GPU's, one 5550' and two 5750's. The 5550 is running oc'd at 700 MHz, and the two 5750's are running oc'd at 860 MHz. I have to have the side panel of the case off and a big house fan pushed up flush with the side of the case blowing on the highest house fan setting and I still can't get them temps down, the 5750's average a temp of 80C. If I turn the house fan off the computer crashes and turns off within 90 seconds. It's in my bedroom to so the house fan is loud. With there was another method of keeping these cards at 80C or below without a big house fan.  ???

None of my rigs motherboards are in cases. Also try laying a 120mm Fan on top of the GPU's once the board is on a flat surface.

Should I change the power to +20%?

At 1150 MHz I'm getting 675MH/s average with GUIMiner

675 is very close. But try the powertune, you have nothing to loose. You can also try to lower the -f parameter (to about 10-15) but your desktop will be (much) less responsive.

I've cut 2 holes in the side panel and put 2x 16cm silent fans there, blowing directly on 2 nvidia fanless 8600gt cards in sli. Went from 100C topcard and 84 bottem one to 82 and 75C. Still hot and I'm only surfing the net with this pc. Occasionally I play trackmania racing and it goes to 120C on the top one. Still not a problem, but it smells like there's a kettle on a fire. And I made a carton wall avoiding cpu heat reaching the vid. cards, a sort of top compartment leading cpu heatflow towards the psu which pulls it out.

I've got a excellent dual GPU's on my gaming laptop, keep it propped up with a extra fan blowing air underneath and it runs at about 70-72 which is pretty cool compared to the heat they are designed to put up with.

My hardware is fully insured otherwise I'd never even consider trying it.

I was mining almost 680 MH/s, and yesterday, for no reason was all day around 290 MH/s

now, my values are:

     Activity: 99%
     GPU clock: 1150%
     Power: 0%
     Temperature: 64%
     Fan Speed: 85%

could it be that it is broken????   already?? I was using it for 4 days only O_O why is not mining at what it's supposed to?

Edit: I used 50BTC, Deepbit and Bitminter. Different times, different programs to mine. It never went over 290 MH/s

Make sure your PC is set to Performance mode not power saving mode...

I will do...thanks!!

My 7970s run at 78°C with 50-55% fan, -20% Power Setting, 640Mh per piece with 1125 (C3D) or 1025 (Sap) MHz.
My 6950 has a new 120mm fan and has 72°C with 100% fan (1500rpm), 320Mh with 840MHz at -20% Power.
All C3D and Sapphire with AMD design.

All in standard cases, but i have no sidewall mounted. Also, one 120mm fan blows cold air on the cards for my twin 7 machines.

70c Seems normal.

When they get to 90c the solder starts to go brittle so worry in the high 80c's

Turn the memory clock down as low as it can go saves power & heat

If you have i5/i7 the motherboard doesn't reach full speed until you're in turbo mode, mines faster when I'm gaming on one card.

I think 70 c is normal but not too low also.

in general many video cards keep their temp at 60º-70ºC,You could fry it if its to hot, but most video cards will just shut them selves off

Coming from the gaming side of things, 70+ is ok but I'm not sure if I'd want to run it 24/7 (with certain cards, such as the HD4850, you had no choice though). 80+ is definitely bad.

As others have said, I would play around a bit with the fan settings if I were you. After a certain point (often 80-85% in my experience) the fans move air with equal/less efficiency and are just generating heat, wasting energy and wearing out more quickly.