Amd R9 290x black screen / reboot issues

Question

Hey there, I recently puechased an alienware aurora r4 that came with an Amd r9 290x and im having los of issues. After playing 45 minutes with a 2 year old game my computer restarted. I ve been checking through the internet and i cant find the solution for this. Dell's tech support sent me another card and did exactly the same. Tried to update to the newest drivers and still the same. Im not able to perform a benchmark without getting a blackscreen / reboot. Can you help me guys? PS: sorry for my bad english

rr1107 · Accepted Answer

Your HWMonitor photos aren't very helpful because you can only see the temperature for the GPU; you might want to try GPU-Z which will show much more information about the GPU including voltages, fan speeds, GPU utilization, clock speed and other important information on the GPU. You could also enable logging in GPU-Z and run a benchmark or game until you get a crash. The log will have the results right up until the computer crashes, and the last entries in the log may show something useful. Also, your logged temp. of 90C is completely normal for this card, so overheating seems unlikely; the card is expected to run up to 94C.

All of that said, it looks as if your GPU is being run on one PCI-E power cable that is "split" into 2 connectors. I can't tell from your photos, but you may have another PCI-E power connector on your power supply. To check, look for a cable coming off of the power supply that looks like the one currently connected to your GPU, but likely with different colored wires. It would be a single bundle of wires with two connectors on the end, one with 6-pins and another with 6+2 pins. If you have the 2nd set of connectors, they may be tucked underneath the HDD trays (bottom-right in your photos). Ideally, if you have 2 of these cables, you would like to use 1 connector from each cable so that you aren't pulling all of the power for the card off of one of the power supply's 12Volt "rails." In the first picture of your case interior (lower left of the picture) it seems that you can see the label on the side of your power supply. A clear photo of this label (the black one) could help determine if the power supply has an additional PCI-E power cable, and also how much power each cable can provide. Unfortunately I am not familiar with the power supply in your particular computer to be able to tell you what connectors your PSU might have.

Also, please clarify again what type of crash you are having. After 45 minutes, does the computer reboot on its own, or does it lock and require you to manually shut it of and restart? This information is important and may help.

This is also all assuming that your problem is in fact the GPU, which it might not necessarily be. Since you are getting the problem with both the old and new GPUs, you may be looking at something else as the cause. As I mentioned before, try switching to a different cable to connect the graphics card to the monitor, or try a different port on the graphics card. You could also try a different monitor if you have one available. These are several very simple steps that you can try to see if they make a difference before you begin more drastic troubleshooting steps or return another GPU as defective.

Sass182 · Answer

Forgot to say the gpu starts a 60 degrees temp on desktop and chrome open and raises to 90-95 when i start playing.

Alienware-Naomi · Answer

Hi, It seems the video card is overheating. Try the troubleshooting steps in this article and let me know the results. If the issue persists, the video card might need another replacement.

rr1107 · Answer

94C is the upper limit for the operational temperatures of the R9 290X, and the "reference" design cards will usually hit that number under load (and then begin to automatically throttle the clock speed), so that seems to me to actually be within the normal thermal expectations for the card. There are several different known black screen issues with the R9 290X (I have had recent experience with black screen crashes and the Asus R9 290X DCII with an upgrade I did to my Area-51) so knowing which type of crash you are having will help.

If the screen is going black and the system "locks" (usually with the audio getting stuck and "looping") and requires a hard reset (ie. holding the power button down and forcing the system to restart), that may be a sign of a bad card. This was the type of black screen crash that I was having. The card would operate fine at the desktop, for web browsing or sitting at the menu screens for a game, but as soon as I would start playing a game (or running a benchmark), the computer would black screen and lock, requiring a hard restart. The original card turned out to be defective and the replacement card has worked perfectly. There are several potential causes for this type of crash, including a faulty DVI/HDMI connection, the GPU overheating, too aggressive of an overclock, various GPU card BIOS issues, faulty GPU memory, or bad voltage regulators. You may also want to try swapping out your DVI/HDMI cable and retest. If you are using a DVI cable, try switching between the 2 DVI ports on the card to see if the problem persists.

If the screen goes black and the system "powers down" and restarts on its own, that may indicate that your power supply is not providing enough power for the card. The R9 290X alone requires around 300Watts (25Amps @ 12Volts) combined from your motherboard and your PCI-E power cables; I believe that AMD recommends a minimum 700W power supply. If the card is being connected with power cables that are supplied by a common PSU "rail" then depending on the power supply and how its PCI-E power cables are set-up, the power supply may be overloaded resulting in a fault protection shutdown and restart.

I hope some of this helps; I was very disappointed when my new R9 290X arrived an wouldn't run any games for more than a few seconds without locking up, however the problem turned out to be a faulty card and the replacement has worked perfectly since.

Good luck!

Sass182 · Answer

Hey there,

Thank you all for your reply's. I have to say got this computer one week ago and im having this issues since the first day. Tech support sent me a card replacement and its doing the same as the one that came with the computer.

RR1107 the issue it's that i cant stand playing a game more than 45 minutes straight without getting a reboot (no BSOD). PSU its an 875w that came with the computer itself. And if i try to run a benchmark (3DMark demo Fire Strike test) it reboots even when loading the test.

I've read Naomi's article and everything seems to be normal (clean of dust, last bios installed A11, last drivers etc...).

Im adding some images of HWMonitor and inside wiring so you can check seems its all "normal (feel free to correct me if im wrong)

Hope you can see values correctly. Only with system startup its at 53 degrees

I just started a game (this is case its TERA) and without doing nothing jumped to 90-92 degrees. If i stress it in a zone with full effects will only stand for 30-45 minutes.

Ok now ill add some inside photos (sorry if its bad quality dont have a better camera) hope will be enough.

Wiring under the card:

image[4].jpeg

Wiring with a different angle:

image[5].jpeg

Before i asked you guys i called tech support and the asked me to do the next steps::

Remove RAMS and run benchmarks (reboot)

Run full ePSA for errors (7 hours of test no errors)

Clean OS install (reboot)

Im really desperate and dont know what to do more.

Thanks for your help.

Sass182.

Sass182 · Answer

Hey RR1107,

I only have one thing to say... YOU ARE AWESOME!

Did what you said, connected the extra PCI cable and it worked perfectly! Could run 3 stress test without getting a single reboot. Seems that the GPU wasnt getting enough power to cool itself.

I think the problem its solved but i still want to test it in a full game session.

Thanks for all. You were really usefull!

Best regards.

rr1107 · Answer

You're welcome and I'm glad to hear that things seem to be working now. What was probably happening was that your GPU, when it was being stressed, was drawing more power than the PSU could provide through that single cable. When too much current is drawn from the power supply, the power supply shuts down and restarts so that the power supply and the system aren't damaged. By using the two different cables, you are splitting the power draw between two of the power supply's internal "rails" and reducing the stress on each individual "rail."

You may want to look at the OCCT stress testing suite and run the "Power Supply" test to make sure everything is nice and stable. It is free and should confirm that everything is ok with your power supply. Otherwise, just play some demanding games and make sure everything runs correctly.

Cass-Ole · Answer

There is a current post titled: 'Aurora R4 - R9 290X Crossfire Performance Issues and System Reboots' by member Alex.e.zieba: http://en.community.dell.com/owners-club/alienware/f/3746/t/19590286.aspx replies are currently disabled on that thread, at least for me. Perhaps Alex can find what I've posted here, since the issue he's having sounds related to this one. Dual 290's? Impressive. Most impressive, (if they'll work). Clik this & follow the chart for proper cable assignments & report back: http://www.dell.com/support/troubleshooting/us/en/19/KCS/KcsArticles/ArticleView?c=us&docid=430542&doclang=en&l=en&s=gen&cs= (in the end, it may be your psu isn't powerful enough for both cards, since the solo app in this thread requires two separate rails to get enough juice)

Cass-Ole · Answer

I'm going to ruminate a minute: as we pan up & look into sass182's case & see his solo pci-e cable, & we're told it by itself can not run a 290x, & so he's been correctly advised to use both his pci-e cables on the one card & that was the proper cure ... it would not surprise me then if the 875w could not run dual 290's, simply for the lack of enough pci-e cables (a lack of 12v rails). They gave you two cables (two 12v rails) & it looks like you need four cables / four rails. When I posted lastnight about correct cable assignments using the Alienware psu cable graph, the graph helps solve a solo card application, but won't bail you out of a dual card app - not cards of that caliber - I don't think

ab1976 · Answer

Thanks for this. I was having the same problem with my 290x (which I bought for my Aurora R4) and resolved it by plugging in cables 15 and 17 off two separate rails into the 290x. One rail is not enough to power it. I am having no more sudden shut downs as a result. Man thanks. Andrew

Locked Topics – Alienware General

Amd R9 290x black screen / reboot issues

Was this post helpful?