Caporegime
.. the general population.
Fixed for accuracy!!!
Hehe. How does yours fare?
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
*** Official AMD 7990 Owners Thread ***
- Thread starter Catanonia
- Start date
Hehe. How does yours fare?
Soldato
Yeah mine maxes out at 1125Mhz with about 1.225v. My memory is decent and can hit 1750Mhz.
I had two 7970 Lightnings and neither could hit 1.2ghz at sensible volts, so I'm not too disappointed.
Soldato
Mate mine is shocking these days..
Caporegime
That's pretty good. Think i need around 1.250v at 1125, though that may not be 100% stable but it did pass a couple of hours of BF4 single player.
Go on...
Hi guys, I made my own thread but thought id post here anyway. I am all of a sudden having major problems with my 7990.
It is in my PC with a 7950, all has been fine since Ive had them together and all of a sudden whilst playing a game on the 7950, (mining on the 7990) the screen froze and I had to reboot.
Upon reboot I went to reprogram it only to find that it will only detect one of the GPU's from the 7990. I have tried everything from reinstalling drivers to swapping PCI-E slots and I cannot fit the problem. In GPU-Z it says crossfire on the 7990 is disabled, and on the catalyst it says : "The AMD CrossFireX internal bridge interconnects linking your graphics cards together are not properly attached. Both bridge interconnects must be attached. For more information, see the user's guide for your graphics card."
I dont really know what else I can do.. I have tried to only boot with just the 7990 in my PC however for some reason the PC wont boot unless the dvi cable is plugged into my 7950, otherwise it will just turn on and be stationary.
Any help would be truly appreciated!!
It is in my PC with a 7950, all has been fine since Ive had them together and all of a sudden whilst playing a game on the 7950, (mining on the 7990) the screen froze and I had to reboot.
Upon reboot I went to reprogram it only to find that it will only detect one of the GPU's from the 7990. I have tried everything from reinstalling drivers to swapping PCI-E slots and I cannot fit the problem. In GPU-Z it says crossfire on the 7990 is disabled, and on the catalyst it says : "The AMD CrossFireX internal bridge interconnects linking your graphics cards together are not properly attached. Both bridge interconnects must be attached. For more information, see the user's guide for your graphics card."
I dont really know what else I can do.. I have tried to only boot with just the 7990 in my PC however for some reason the PC wont boot unless the dvi cable is plugged into my 7950, otherwise it will just turn on and be stationary.
Any help would be truly appreciated!!
Man of Honour
- Joined
- 21 Nov 2004
- Posts
- 45,776
The internal bridge? That doesn't sound good at all.
Soldato
Ouch, sounds like it's goosed. One of my biggest fears I'm afraid when it comes to mining - so much so I backed off the hotter core mining at a lower intensity than the cooler core so that temperatures remained steady.
I found AB or cgminer would not hold the under-volt figure and would catch my hot core into the 90's before shutting it down and re-applying the UV.
I was wondering if any other users had their cores mining with differing characteristics (settings)?
I found AB or cgminer would not hold the under-volt figure and would catch my hot core into the 90's before shutting it down and re-applying the UV.
I was wondering if any other users had their cores mining with differing characteristics (settings)?
Associate
I grabbed one during the Black Friday sales for £360 and I've got to say this thing is truly amazing. I need to learn how to overclock though as my i7 965 is running stock @ 3.2Ghz but it will run anything I play except far cry 3 and Arma 3 (both cpu intensive) maxed out at around 100fps+ Will post some pictures soon 
Associate
Just thought I would update everyone of my plight with the 7990.
Stability wise, it's holding it's own, I have the odd lockup under load, but it has reduced considerably. I tried replacing the thermal paste on the gpu's to alleviate the temperature differences and it made zero difference (but it has coincided with the card acting more stable). Now that Afterburner can show stats in 64bit applications, I did a session of 3dmark and it got scary.
Click to see image, too large to place on here
It appears to get a fair few degree's hotter in 3dmark than it does in Heaven.
RMA hasn't gone tremendously well so far. After receiving my RMA number, I had to send an additional note to the RMA to ensure they got all the information recorded that they would need, which was not responded to for 2 weeks. Another message to them was then swiftly answered, but as it's so close to the christmas period and I don't want to be without a card over the christmas holidays. I will persevere with the problems until Christmas is done with and will then send it back.
As I said, changing the thermal paste has made no noticable differences to temperature in the slightest. The general stabillity of the card has improved tremendously though and it's no longer the headache it was. It's still annoying that I have to deal with it having spent £480 on it, but it's a lot more manageable now.
The main issue is definitely the temperature though. My concern is, these are the temps in the winter, in the summer it will be well beyond any means of controllability as I am in a room with no source of heating, so the ambient temperature is low. If you add 10+ degree's to the ambient temperature, there is no way on this earth I will be able to control the temperature. With a 29c difference in temps between gpu1 and gpu2, something has to be up with it, but I can't put my finger on it.
For reference, I have never seen gpu2 rise above 72c.
So for now, I am all about limiting fps in games as much as I can to try and keep the temperature under control. Not ideal, but it's better than having to use the onboard haswell chip over the christmas holidays. Hopefully I can get it all sorted after christmas and get to the worry free gaming, which is why I actually put this much money into a system for in the first place.
Sorry for the long post, you will be thankful when I can put this all to some (hopefully) satisfying conclusion, so you don't have to read my ramblings about it.
Stability wise, it's holding it's own, I have the odd lockup under load, but it has reduced considerably. I tried replacing the thermal paste on the gpu's to alleviate the temperature differences and it made zero difference (but it has coincided with the card acting more stable). Now that Afterburner can show stats in 64bit applications, I did a session of 3dmark and it got scary.
Click to see image, too large to place on here
It appears to get a fair few degree's hotter in 3dmark than it does in Heaven.
RMA hasn't gone tremendously well so far. After receiving my RMA number, I had to send an additional note to the RMA to ensure they got all the information recorded that they would need, which was not responded to for 2 weeks. Another message to them was then swiftly answered, but as it's so close to the christmas period and I don't want to be without a card over the christmas holidays. I will persevere with the problems until Christmas is done with and will then send it back.
As I said, changing the thermal paste has made no noticable differences to temperature in the slightest. The general stabillity of the card has improved tremendously though and it's no longer the headache it was. It's still annoying that I have to deal with it having spent £480 on it, but it's a lot more manageable now.
The main issue is definitely the temperature though. My concern is, these are the temps in the winter, in the summer it will be well beyond any means of controllability as I am in a room with no source of heating, so the ambient temperature is low. If you add 10+ degree's to the ambient temperature, there is no way on this earth I will be able to control the temperature. With a 29c difference in temps between gpu1 and gpu2, something has to be up with it, but I can't put my finger on it.
For reference, I have never seen gpu2 rise above 72c.
So for now, I am all about limiting fps in games as much as I can to try and keep the temperature under control. Not ideal, but it's better than having to use the onboard haswell chip over the christmas holidays. Hopefully I can get it all sorted after christmas and get to the worry free gaming, which is why I actually put this much money into a system for in the first place.
Sorry for the long post, you will be thankful when I can put this all to some (hopefully) satisfying conclusion, so you don't have to read my ramblings about it.
Caporegime
Something is definitely wrong with your card mate. No way it should ne anywhere near 100c at stock clocks/voltage. Pretty sure if i ran it and matched your settings my temps would be much lower.
Soldato
Interesting Buz, thanks for sharing. I too have noticed the core inconsistency. At the beginning they were only 2 or 3c difference, but I suppose mining will highlight any flaws as it's getting used constantly.
In games it has been fine but that would be due to the games not stressing the GPU/CPU to the max.
I could easily thrash this thing but now I know it can go into the 90s unattended I dont think it would last too long operating at that temperature. Long term if I could afford it I would not hesitate and try get it under water or use an AIO option.
My case has fantastic airflow, and sporadic gaming use would pose no problem. Pushing these cards for heavy use certainly changes the beast. I was only intending to do the mining for the short term while there was money to be made. If the additional heating was not required I would stop mining shortly, but as it tops up the downstairs nicely saving me on oil heating the mrs actually likes it.
Interesting about the thermal compound, I thought you would have seen a shaving there of 2c. As for the second core mine too runs about 72c full load.
What make is yours btw?
In games it has been fine but that would be due to the games not stressing the GPU/CPU to the max.
I could easily thrash this thing but now I know it can go into the 90s unattended I dont think it would last too long operating at that temperature. Long term if I could afford it I would not hesitate and try get it under water or use an AIO option.
My case has fantastic airflow, and sporadic gaming use would pose no problem. Pushing these cards for heavy use certainly changes the beast. I was only intending to do the mining for the short term while there was money to be made. If the additional heating was not required I would stop mining shortly, but as it tops up the downstairs nicely saving me on oil heating the mrs actually likes it.
Interesting about the thermal compound, I thought you would have seen a shaving there of 2c. As for the second core mine too runs about 72c full load.
What make is yours btw?
Associate
It's Gigabyte.
I was surprised about the thermal paste not affecting things. If not positively, then surely negatively, for temperatures to stay the same it's not really heared of.. It literally hasn't made a noticeable difference. Especially GPU2, which doesn't budge from that 72c under load, since before or after the change.
When I opened it up, I was certain the thermal compound would improve it as the original paste looked slopped on haphazardly. It was all overflowing over the sides and onto the board and it generally looked very messy.
It made no difference at all though, 3d load improved tremendously. I can't explain how that would happen from the thermal paste, when temps never changed. So I will put it down to coincidence, but you never know. The problem is, I know at the current rate I can't control this card when ambient temperature rises. I could put it under water, but you are talking a couple of hundred pound to get results that I should realistically be seeing from air anyway. If it is faulty, then it's potentially a waste of time and money also.
VRM temps are generally well under control also I must add, it's purely gpu1 that causes the problems and gpu1 only.
As I said, hopefully after christmas the RMA will be sorted and it will come to a satisfying conclusion.
I was surprised about the thermal paste not affecting things. If not positively, then surely negatively, for temperatures to stay the same it's not really heared of.. It literally hasn't made a noticeable difference. Especially GPU2, which doesn't budge from that 72c under load, since before or after the change.
When I opened it up, I was certain the thermal compound would improve it as the original paste looked slopped on haphazardly. It was all overflowing over the sides and onto the board and it generally looked very messy.
It made no difference at all though, 3d load improved tremendously. I can't explain how that would happen from the thermal paste, when temps never changed. So I will put it down to coincidence, but you never know. The problem is, I know at the current rate I can't control this card when ambient temperature rises. I could put it under water, but you are talking a couple of hundred pound to get results that I should realistically be seeing from air anyway. If it is faulty, then it's potentially a waste of time and money also.
VRM temps are generally well under control also I must add, it's purely gpu1 that causes the problems and gpu1 only.
As I said, hopefully after christmas the RMA will be sorted and it will come to a satisfying conclusion.
Caporegime
Silly question but have you checked to make sure all the fans are spinning?
Associate
They are definitely spinning.
If gpu2 was also going to much higher temps, I would probably be concerned about my cooling solutions. That is easily under control though and is around the temps I would probably expect from my setup. I have the Cooler Master HAF 932 Advanced Full Tower case, with 230mm front, 230mm side, 230mm top and 140mm rear fan setup, which I tried in many different intake/exhaust configurations before settling with something more akin to the setup you posted from you case a few weeks back.
If gpu2 was also going to much higher temps, I would probably be concerned about my cooling solutions. That is easily under control though and is around the temps I would probably expect from my setup. I have the Cooler Master HAF 932 Advanced Full Tower case, with 230mm front, 230mm side, 230mm top and 140mm rear fan setup, which I tried in many different intake/exhaust configurations before settling with something more akin to the setup you posted from you case a few weeks back.
Associate
To give an example of the cooling performance inside the case. I have the 4670k at 4.6ghz with 1.3v voltage, with the Alpenphonn Mount Doom cooling it and the temps are very good staying in the 50-60c range whilst playing BF4. I could probably throw more voltage and clock at it and still have it manageable, but don't really feel the need to increase it any more.
The airflow is working wonders for everything except that one damn gpu!
The airflow is working wonders for everything except that one damn gpu!
Associate
If you want to see the true uncontrollable nature of it, I throw the fans to constantly run at 100% and then start stressing the card. The below images are what you will see. Crazy right?
Clicky
39c difference!
Clicky
39c difference!
Caporegime
Yeah definitely something not right there. 10-15C difference is within tolerances but any higher than that and there is something wrong. I'm lucky that my temps are always identical or within 1c of each other.
Associate
Yeah, well hopefully after christmas I will sort it out. If they agree that this isn't right. Then I am unsure I will be able to get a replacement 7990 due to recent shortages, which is kind of unfortunate, because apart from the fundamental flaws I have had, the card is an absolute beast and definitely has a place in my system.