Sign in with
Sign up | Sign in
Your question

R9 290 crashes and impossible GPU wattage

Tags:
  • GPUs
  • PowerColor
  • Graphics
Last response: in Graphics & Displays
Share
August 6, 2014 9:15:47 AM

Hi,

so I just got a new rig, which consists of a second-hand Powercolor R9 290.

Specs:
-AMD Phenom II X6 1075T
-Asus M4A87TD
-6GB RAM
-Powercolor R9 290
-Corsair HX850 PSU

It proved to be pretty good as I could run quite a few games on full/maxed-out settings (Borderlands 2, Deus Ex: HR, Dota 2) with good performance (didn't run a framerate test but I did not notice any stuttering)

However, the rig would crash sometime into the game (ranging from 5min-30min) and would often require a hard restart. Sometimes it would crash to a black screen of death, other times it would come up with lots of static on screen.

Temperatures get pretty high sometimes (80-90C) but I've read online that that's ok because AMD says they can go up to 95C. I don't think it's an issue with temp as it runs really smoothly up to the point where it crashes and doesn't show any noticeable framerate throttling, which would be the case if heat was the issue.

I'm using AMD Catalyst Control Center 13.11, as it crashed more often with the up-to-date driver. Uninstalled 14.4 using Display Driver Uninstaller and installed driver from the disk provided with the GPU

Tried using GPU-Z to see if anything was amiss and I found that VDDC Power in/out was approx 8W during idle, with max value of 40W under load.
I know this definitely isn't normal as desktop GPUs usually use about 200-300W and this isn't even remotely close.

Help me please! I'm at my wits end trying to figure this out. Would RMA'ing help?

More about : 290 crashes impossible gpu wattage

August 6, 2014 9:52:19 AM

Yes! RMA the card. Artifacting (or static) is a good sign it's a faulty card.
It sucks that someone sold it off to you. I bet it's the reason they did so. :( 
Best of luck!
m
0
l
August 6, 2014 9:52:40 AM

When you installed the card, did you plug in the extra power connectors firmly so that they snapped into place?

And no, these temps aren't normal, only for a stock R9 290X, otherwise you should get somewhere in the 70's or 80's, but even those are too much. I have a sapphire R9 280X with the shitty dual-x cooler, even with that one, the temps don't get above 65°C!
m
0
l
Related resources
August 6, 2014 10:51:31 AM

As above, used card, need to test it before buying. Many people sell cards that crash like this unortunately. If you can't test a used card, don't get it. Unless it's from a reputable place that will protect you from issues with returning to the seller.

What do you mean by RMA exactly? If it's used, the vendor may not take it back unless you are the original owner with a valid reciept. I know the "lifetime" waranty of a certain vendor is only the lifetime of the original owner having it. Once it's passed to another person, warranty is done.

The last owner could have tried to overclock it a bit too much and damaged it, this also would not be covered under warranty.
m
0
l
August 6, 2014 10:56:13 AM

hang-the-9 said:
As above, used card, need to test it before buying. Many people sell cards that crash like this unortunately. If you can't test a used card, don't get it. Unless it's from a reputable place that will protect you from issues with returning to the seller.

What do you mean by RMA exactly? If it's used, the vendor may not take it back unless you are the original owner with a valid reciept. I know the "lifetime" waranty of a certain vendor is only the lifetime of the original owner having it. Once it's passed to another person, warranty is done.

The last owner could have tried to overclock it a bit too much and damaged it, this also would not be covered under warranty.


I completely agree with this. If he can get in contact with the seller (which he probably won't be able to considering they were horrible enough to sell him a broken card) he should be able to RMA it, if not it's pretty much done. :( 
m
0
l
August 6, 2014 7:03:41 PM

Hi all, thanks for all your replies!

I live in Singapore, and there's a sole distributor in my country that manages all the RMAs, regardless of where I actually bought the card from. In my case I got the pre-owned card from a listing in the classifieds and the seller provided me with the receipt and serial number.

He seemed trustworthy enough when I met him and when I told him I was having issues with the card, he did try to help. (Tried underclocking and increasing fan speed using MSI AB to decrease temps)

When it was apparent that it wasn't working, he told me to RMA the card at the distributor outlet. He said the place would accept the RMA request as long as I have the serial no.

In the event that the previous owner did overclock it, is there any way for the distributor to find out? And if so, is the warranty void for sure?
m
0
l
August 6, 2014 7:04:48 PM

Good to hear! The warranty will be fine. OCing doesn't void it. :) 
m
0
l
August 6, 2014 7:11:31 PM

BigBadBeef said:
When you installed the card, did you plug in the extra power connectors firmly so that they snapped into place?

And no, these temps aren't normal, only for a stock R9 290X, otherwise you should get somewhere in the 70's or 80's, but even those are too much. I have a sapphire R9 280X with the shitty dual-x cooler, even with that one, the temps don't get above 65°C!


What do you mean by extra power connectors? I did plug in the 6-pin and 8-pin PCI-E connectors from the PSU if that's what you mean.

Anyways it gets pretty hot where I live (25-30C room temp.) so I've been considering getting a liquid cooler for the GPU. Any recommendations on good and affordable ones?
m
0
l
August 6, 2014 7:22:42 PM

Alex Kelly said:
Good to hear! The warranty will be fine. OCing doesn't void it. :) 


That's great news :)  will send it for RMA this weekend and update on the situation. That being said, I'm kind of worried that it might also be a problem with the PSU, judging from the GPU power draw and the fact that the GPU looks slightly undervolted (normal voltage is ~1.15 right?)

Are VDDC power draw, voltage and current controlled by the GPU or should I be concerned with the state of my PSU?
m
0
l
!