• 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.

Help needed - R9 290x Issues

Gadget1st said:
Thanks for the suggestion, i'll have a play with the settings tonight when I'm home from work.

Andrew
Click to expand...

I installed Afterburner and here are the results.

Memory Clock set to 1100.

I firstly used a graduating fan speed and this got the GPU up to 97C and then World of Tanks froze, I did CTRL+ALT+DEL and got into Task Manager and closed down WoT.

I then set the fan speed to 56% and tried WoT again and the GPU got to 59C and WoT froze again.

In both instances when I closed WoT there were sparkles on my desktop, only after the GPU cooled down enough did the sparkles disappear.

I look forward to your next suggestion after you have taken a look at the log from AfetrBurner.

Many thanks
Andrew
 
I've read that memory clocks that 3rd party non-reference cards were given with insufficient ram voltage cause a lot of problems. It also might be the case that this card has been used as a bit-coin miner, so 24/7 at the limit. Then people found that after that the cards were unstable.

I've looked at your results (in riva tuner) and it doesn't give me any different advice to what has already been posted.

If you've changed the thermal paste, downclocked the memory, then next simplest thing would be to bump the gpu voltage.

However in an attempt to try to attribute the instability to either the gpu or ram. You could try further downclocking the ram or keep the ram the same and downclock the gpu.

I think that testing with 100% fan speed could be helpful before you try anymore downclocking.

Another thing is - I'm not seeing the VRM temps, the "aux voltage" maybe the amplitude of the output of the VRM but it would be remiss to ignore VRM altogether.

If it had a warranty I would suggest RMA'ing it but at this point you probably want to try anything to make it behave.

You should be able to redefine the graph's limits so that you don't get exclamation marks next to the graph's range, some limits are off the chart, other ranges are too broad.

Unfortunately there is no way to increase memory voltage at least without a soldering iron.
 
Last edited:
megatron said:
I've read that memory clocks that 3rd party non-reference cards were given with insufficient ram voltage cause a lot of problems. It also might be the case that this card has been used as a bit-coin miner, so 24/7 at the limit. Then people found that after that the cards were unstable.

I've looked at your results (in riva tuner) and it doesn't give me any different advice to what has already been posted.

If you've changed the thermal paste, downclocked the memory, then next simplest thing would be to bump the gpu voltage.

However in an attempt to try to attribute the instability to either the gpu or ram. You could try further downclocking the ram or keep the ram the same and downclock the gpu.

I think that testing with 100% fan speed could be helpful before you try anymore downclocking.

Another thing is - I'm not seeing the VRM temps, the "aux voltage" maybe the amplitude of the output of the VRM but it would be remiss to ignore VRM altogether.

If it had a warranty I would suggest RMA'ing it but at this point you probably want to try anything to make it behave.

You should be able to redefine the graph's limits so that you don't get exclamation marks next to the graph's range, some limits are off the chart, other ranges are too broad.

Unfortunately there is no way to increase memory voltage at least without a soldier iron.
Click to expand...

Here is the latest, the result @ 21:56:58 is the one under stress:

Fan @ 100% and playing WoT @ max settings getting 60fps, where it drops off is when WoT crashed and I CTRL+ALT+DEL to close it although no sparkles on screen when I got back to desktop.

Thanks again, I feel we are getting somewhere :)

Andrew

p.s. Just upped the Memory to 1150 and when I hit apply it went black screen, coincidence?
 
Last edited:
According to the 21:56 log you are running @ 1250Mhz GDDR5 (x 4 effective 5000Mhz) and between 850 and 950 Mhz gpu. The specifications are 1000gpu 5000 GDDR5.
http://www.sapphiretech.com/presentation/product/product_index.aspx?pid=2060&lid=1#

So you are running full speed on ram and unless you are intentionally underclocking the GPU then its throttling.

Its also not clear at what precise time you crashed in WoT. That might help get a general idea of what is going on.

The black screens are I believe attributed to the ram and sparklies are definitely an indication that the RAM is unstable.

Your GPU is 59 degrees tops in that 21:56 log and you say game crashed, while the gpu was underclocking to 850-950.


If I was you I would keep the GPU settings and set the RAM to 1075Mhz (4300Mhz effective) since you already said that 1100Mhz had crashed on you (correct me if I'm wrong). Also you could try directing a fan at the RAM.

What is your impression? Where you able to plan WoT for a longer period before instability? Also write down the log, the time you become unstable, the GPU/RAM clocks in addition to posting the log so that I can verify what I see in the log is the clock speeds you were intending to test with.
 
Last edited:
megatron said:
According to the 21:56 log you are running @ 1250Mhz GDDR5 (x 4 effective 5000Mhz) and between 850 and 950 Mhz gpu. The specifications are 1000gpu 5000 GDDR5.
http://www.sapphiretech.com/presentation/product/product_index.aspx?pid=2060&lid=1#

So you are running full speed on ram and unless you are intentionally underclocking the GPU then its throttling.

Its also not clear at what precise time you crashed in WoT. That might help get a general idea of what is going on.

The black screens are I believe attributed to the ram and sparklies are definitely an indication that the RAM is unstable.

Your GPU is 59 degrees tops in that 21:56 log and you say game crashed, while the gpu was underclocking to 850-950.


If I was you I would keep the GPU settings and set the RAM to 1075Mhz (4300Mhz effective) since you already said that 1100Mhz had crashed on you (correct me if I'm wrong). Also you could try directing a fan at the RAM.

What is your impression? Where you able to plan WoT for a longer period before instability? Also write down the log, the time you become unstable, the GPU/RAM clocks in addition to posting the log so that I can verify what I see in the log is the clock speeds you were intending to test with.
Click to expand...

Thanks Megatron

I will try the settings you suggest when I get home tonight.

Once again I would like to thank all the members who have helped me along the way and hopefully this thread will help others who have the same issue and come to a satifactory conclusion.

Many thanks
Andrew
 
Deffo RAM instability, shame because these 290x's are beasts.

I use to benchmark these cards when they came out and found that core voltage helped with memory instability a bit. Would be worth putting power to +20%, keeping core the same, adding maybe 25mv to the core voltage and turning the memory down to 1050/75.
 
Good morning guys

Here are the latest results from last night.

RAM set to 1075, upped the power by 20 and I did add 25mv to the core.

The card was stable (in game) albeit a bit noisy as the fan was set to increase on a curve adjacent to the heat.

Where the graph ends is where the game crashed and went black screen, I could hear game noises but it was the same noise not changing when keys were pressed to move forward or backward. I could not CTRL+ALT+DEL and had to press power button for 4+ secs to power down.

I have no way of pointing a fan at the ram as this is a reference card and has a cover over heatsink to force cool air over the fins, if I remove the cover the air is not cooling the heatsink.

I really appreciate all your suggestions and hope that this issue is either resolved or I save up for a new card. I can still play WoT with my R7 240 but to get a decent framerate I have to really downgrade the graphics and it doesn't look great (in comparison to the r9 290x). I know now I should have saved a little more money and bought new but the playability of the game was an over riding factor.

Thanks again.
Andrew
 
The VRAM is fine temperature wise. Is there any way you can contact the seller? If you have to lower the VRAM clocks too much just to get it working, you have a pretty crippled card and it would definitely be worth pushing for a return if it was bought on the bay or on a second hand moderated forum.
 
New Card then?

Avenged7Fold said:
The VRAM is fine temperature wise. Is there any way you can contact the seller? If you have to lower the VRAM clocks too much just to get it working, you have a pretty crippled card and it would definitely be worth pushing for a return if it was bought on the bay or on a second hand moderated forum.
Click to expand...

Okay I have started proceedings to get my money back so if that works I will have £250+ to spend on a card, what are your recommendations? :confused:

I was thinking: https://www.overclockers.co.uk/showproduct.php?prodid=GX-001-GX&groupid=701&catid=1914&subcat=1010

Thanks
Andrew
 
Last edited:
Atm I kinda regret getting a card that doesn't have a reference style PCB from nvidia. Really wanted to try out one of these AIO things but can't.
 
It sounds like you've skimped out a second time on the psu, which may be the cause of the issue.

You might have better luck with a 970, which uses less power for similar performance.
 
That gtx970 is a good card. I have one. I think it is a bit loud though. Once the fan is at 50% it is the loudest thing in my case. Also it's not got a reference pcb so the corsair hg10 n970 won't fit on it if you later decide you want it. Plus it's also not the acx2.0+. The + has a better cooler apparently.

I mentioned the +. It may be the rev2.00 that has the better cooler. I'm not sure now. Pretty sure I have seen both mentioned.
 
Last edited:
The 290x at those prices. Direct CU cooler keeps it cool and quiet. Even at the same price i would probably go for the 290x in case you decide to go to those higher resolutions.
 
You say you bought a 750w psu but don't say what make and model it is. Not all psu's are created equal and the 290x needs a decent 12v rail. If it's the psu at fault you will get exactly the same problem with a different card.
 
You must log in or register to reply here.
Back
Top Bottom