Sign in with
Sign up | Sign in
Your question

Grey screen crash

Tags:
  • Radeon
  • Crash
  • Graphics
  • Product
Last response: in Graphics & Displays
Share
November 17, 2012 8:28:44 AM

Hi There,

I'm having issues with my PC which cause games to crash to a grey screen with vertical dark lines. On occasion I experience black screen but in most cases I can altf4 or task manager out of it.

The grey screen crash is only stopped by hard reset which can't be good right? I'm at my wits end. I've tried driver clean installs, removal of antivirus as well as performed a number of stress tests and burn in which have reported absolutely no issues. In fact, the core temp is lucky to climb above 45c and GPU around 60c max. No issues with HDD's and no other problems with many games or anything else.

Kicked in after owning the pc for about 2 months and 20 hours on Sleeping Dogs at near max settings. Accumulative.

I'm not very knowledgeable with overclocking or anything like that but if the following information is of any use in hopefully finding a solution I'd be forever grateful:


SPECS:
Windows 7 Home Premium Edition SP1 build 7601 (64-bit)

1 x AMD FX(tm)-8150 Eight-Core Processor [3902.3 MHz]

15.9GB RAM

AMD Radeon HD 7850

2 x 1863GB HDD

CD/DVDRW/BD, CD/DVDRW/BD, CD/DVDRW/BD, CD-RW/DVDRW/BD

MB - M5A99X EVO


CPU features: MMX SSE SSE2 SSE3 SSSE3 SSE4.1 SSE4.2 SSE4a DEP PAE AMD64 AES Turbo core
Clock frequencies:
- Measured CPU speed: 3902.3 MHz [Turbo core: 3902.5MHz]
- Multiplier: x18.0 [Turbo core: x18.0]
- Reference Clock: 216.8 MHz


Kind Regards,

W.

More about : grey screen crash

November 17, 2012 11:29:08 AM

It might be that your card is faulty, can you try it in another computer?
m
0
l
December 30, 2012 3:54:09 AM

According to AMD the problem is with software. It was supposedly an issue with the catalyst drivers but was fixed in updated versions. Yet it seems that the problem still exists in the updated drivers

I'm having similar issues with my XFX Double D Radeon HD 7850 2GB GPU. I've only gotten the grey screen once while gaming. I got the Grey and black pinstripe screen playing ARMA II multiplayer on max settings. I run games like ARMA II, Sleeping Dogs, and Crysis 2 at max settings with little to no problems.

For me, the most common program it happens on is Photoshop. When I first started using it on my computer it would go to the grey screen within seconds of starting a project. I went into the preferences and changed the OpenGL settings to Normal instead of Advanced and it helped out. I am still having issues with the grey screen popping up randomly during sessions though. When I turned off OpenGL entirely it grey screened again (interesting?). I have been using GPU-Z to monitor the GPU's vitals while in Photoshop but haven't seen anything that stands out as a reason for crashing. Perhaps, there is some sort of critical error in the software that makes this happen as opposed to a hardware issue.

Could it also be bad communication between the monitor and the GPU?

Like you I'm getting tired of having to do a hard restart every time this happens. It can create some long term issues. I'm thinking nvidia for my next GPU

I will continue to look for remedies and update if I find one. I fear that an RMA may be in my future though.

My Specs:
AOC 21.5" LED 5ms 1920x1080 @ 60hz
Windows 7 x64
Gigabyte GA-Z77X-D3H LGA 1155 MOBO
Intel i5 3.4GHz CPU
XFX Double D Radeon HD 7850 2GB GPU
G.SKILL Ripjaws X Series 16GB (2 x 8GB) DDR3 @1600 Memory
Western Digital Blue 500GB HD @7200rpm
CORSAIR CX Series CX750 750W ATX12V Power Supply


m
0
l
Related resources
January 7, 2013 10:13:23 PM

Thanks for the information dietdoom. I'm having the same problem that you are experiencing, only with my Radeon HD 5870. I'm running AMD Catalyst 12.10 as well as AMD Overdrive. Windows is also updated.

Like you, I also updated my card fairly recently, only in my case I bought the 5870 used off ebay. So I too am contemplating a return, even though I really would rather just get everything working correctly. I will also update with any new information as I continue searching around and experimenting with solutions.

My specs:
Apple Cinema Display A1081 20" 1680 x 1050 (60 Hz)
Windows 7 Home Premium x64
ASUS M4A785TD-V EVO Mobo
AMD Phenom II X4 965 @ 3.6 GHZ (oc)
Cooler Master Hyper 212 Evo HSF
AMD Radeon HD 5870
10 GB DDR3 @ 1333 (8GB G.SKILL (2 x 4GB) DDR3 @1333 + 2 GB KVR (2 x 1GB) DDR3 @ 1333)
Seagate Barracuda 500GB @ 7200 RPM
OCZ SSD 60GB (OS installed here)
OCZ ModXStream Pro 600W Modular power supply

m
0
l
January 8, 2013 1:36:08 AM

Ok, I switched my drivers to the new Catalyst 12.11 beta. I'm going to try to reproduce the problem now. I'll post again with the results.
m
0
l
January 8, 2013 2:16:42 AM

I also switched to the new 12.11 beta. It seems to be a bit more stable. Haven't had a crash yet (1 week).
m
0
l
January 8, 2013 5:40:34 AM

Still crashing with the 12.11 beta. I also tried switching my memory from unganged to ganged in the bios and removing 2 sticks of ram to bring my memory down to 8 GB DDR3 (2 x 4GB) which is more "normal" than what I had before (10 GB 2x4 2x1). But so far no dice.

Have you guys seen this post? http://www.tomshardware.com/news/ATI-Radeon-Gray-Screen... The quoted stuff from "jogob9" is gold.

I'm going to try raising the voltage now to the GPU and see if that has any effect. Will report back.

EDIT: Oh and I'm absolutely certain that the problem isn't overheating--I had GPUZ log my GPU temps, and right before the crash my last recorded temp was 71C.
m
0
l
January 8, 2013 8:28:34 PM

EDIT: Removing my post so as not to mislead others--I thought I found a solution, but didn't actually.
m
0
l
January 9, 2013 5:39:07 PM

Congrats! That brings up a few questions. Do you get screen flashes and random landscape flickering when you are playing Skyrim? I used to get that bad with the 12.10 drivers but fixed it with the 12.11. Also, how is the screen jitter in first person?
m
0
l
January 9, 2013 6:00:10 PM

I will have to check on the screen jitter in first person since I don't play in first person, but there have been no screen flashes for sure. As for the landscape flickering, if there is any, it is pretty minimal.
m
0
l
January 9, 2013 6:08:18 PM

Nope, no first person jitter for me.

I do want to acknowledge though that this is obviously a sub-optimal solution as it effectively renders the user hostage to obsolete drivers for the duration of the card's life. In my case however it is really my only option, as I cannot RMA the card since I bought it used on ebay. (Going to think twice before doing that again next time, heh).
m
0
l
January 12, 2013 12:15:13 AM

Just a quick update: while I found that my system seemed more stable on the old drivers for a while, I did eventually experience another crash. =[ In a way it's just as well, since I was also dissatisfied with remaining on old drivers as a long term solution anyway.

So I decided to look around on other forums for some alternative solutions. I found a thread on anandtech which suggested flashing the video card's bios to a version with higher voltages. Here's the thread:

http://forums.anandtech.com/showthread.php?t=2219247

The vbios collection that is discussed is found here:

http://www.techpowerup.com/vgabios/

I flashed my card's bios to a version with higher voltages for the four clock states, and I'm back to running the current drivers. So far everything has been stable and I have not experienced a crash. I will report back if I do experience further issues however.

WARNING: if you do flash your video card's bios, be EXTREMELY careful that you select the proper manufacturer. The first time I did this I didn't know that I couldn't use an asus bios for my reference hd 5870, and I got a black screen. It was a major pain to reflash it to the original bios, which brings me to another important point: ALWAYS back up your original bios!
m
0
l
January 12, 2013 10:56:03 AM

I believe I have finally resolved the GSOD on my system.

Win 7 - 64
i7 960
2 x Gigabyte HD 5870 in Crossfire

I underclocked the GPU clock to 750 MHz (default is 850) and the Memory clock to 1100 MHz (default 1200).
Since doing this I have not had any GSOD which I would get within minutes of playing Left for Dead 2 and BF2 Bad Company and I can now play for as long as I want without problem. It has been over a week since I did this.

I have not flashed the card(s) bios so it is still on the original version and the Catalyst drivers in use are 12.4. I have had GSOD
with Catalyst drivers all the way back to 10.9.

I know I have majorly underclocked the cards but my PC it is stable now. I may try incremental increases in MHz
but to be honest I have had the GSOD for so long, now it is gone I can't be bothered.

Anyway maybe for some of you it is worth a try.
m
0
l
January 12, 2013 5:38:57 PM

Thanks for the suggestion. I may try that if I have any more trouble. Did you make a new bios in RBE or did you use some program like afterburner or ccc?
m
0
l
January 12, 2013 5:51:11 PM

Well, I got another crash, so I think I'm going to try your suggestion ratsak.

Do you think the fact that underclocking the card solved your problem means that our issue really is just heat after all? I did a test earlier when I played through a part of skyrim with necromancers that always crashed my system before on the default high settings...I switched the settings to low and it was fine.

High temperatures was one of the first culprits that I thought of when I started getting these crashes and one of the first things that I ruled out, since my readings in GPU-Z were always well within the norms (always under 80c in load). Now I'm thinking though that maybe while the gpu core temperature is fine, some of the other chips on the side may be getting too hot? And that may be the cause of the crash? I purchased my card refurbished, and I opened it up the other day and many of the thermal pads were flaking and cracked and coming off. Now I'm thinking that proper cooling may solve a lot of my problems.

I've ordered this thing from newegg: http://www.newegg.com/Product/Product.aspx?Item=N82E168...

I'll let you guys know how it works out for me.
m
0
l
January 12, 2013 10:21:55 PM

mrxenyo said:
Thanks for the suggestion. I may try that if I have any more trouble. Did you make a new bios in RBE or did you use some program like afterburner or ccc?



I just underclocked the cards using CCC. The cards are using the original BIOS.
I have not altered card voltages at all.

m
0
l
January 12, 2013 10:31:53 PM

mrxenyo said:
Well, I got another crash, so I think I'm going to try your suggestion ratsak.

Do you think the fact that underclocking the card solved your problem means that our issue really is just heat after all? I did a test earlier when I played through a part of skyrim with necromancers that always crashed my system before on the default high settings...I switched the settings to low and it was fine.

High temperatures was one of the first culprits that I thought of when I started getting these crashes and one of the first things that I ruled out, since my readings in GPU-Z were always well within the norms (always under 80c in load). Now I'm thinking though that maybe while the gpu core temperature is fine, some of the other chips on the side may be getting too hot? And that may be the cause of the crash? I purchased my card refurbished, and I opened it up the other day and many of the thermal pads were flaking and cracked and coming off. Now I'm thinking that proper cooling may solve a lot of my problems.

I've ordered this thing from newegg: http://www.newegg.com/Product/Product.aspx?Item=N82E168...

I'll let you guys know how it works out for me.


I thought it was a heat problem too but I had Afterburner Kombuster running for 2 hours at both my default and underclocked speeds. The temperature gets to 87'C without any problem. So I don't think heat is the cause. The stability of the cards at lower MHz suggests to me that the default settings for the cards are wrong for my system. Why Kombuster has no problems on default card settings but games do is a real puzzle?

I haven't noticed any underperformance by underclocking the cards so the only thing I have lost is GSOD.
m
0
l
January 13, 2013 6:49:43 AM

Hey ratsak, so I tried underclocking my card to 750/1100, but whenever I do that I get this screen flickering and tearing that ultimately results in a crash. Is there something else you did to make your system stable that I'm not doing? (I'm back on the original bios that came with the card now--version 012.014.000.005 part# 113-C00101-107 date 2010/09/21 according to ccc).
m
0
l
January 13, 2013 10:36:31 AM

Hi I'm pretty sure everything on CCC is default.
What are the default clockspeeds for your card?
I just underclocked 100 for each and was ok.
m
0
l
January 13, 2013 6:32:09 PM

Default is 850/1200, same as yours.

It just occurs to me that I was technically running a dual monitor setup when I was testing since my hdtv was plugged into the hdmi port at the time, so maybe that had something to do with it? I'll try testing it with the tv unplugged and report back.
m
0
l
January 13, 2013 9:53:35 PM

mrxenyo said:
Default is 850/1200, same as yours.

It just occurs to me that I was technically running a dual monitor setup when I was testing since my hdtv was plugged into the hdmi port at the time, so maybe that had something to do with it? I'll try testing it with the tv unplugged and report back.


According to other posts on screen tearing: If you are still getting screen tearing, you might look at turning "wait for Vertical refresh" and "Triple buffering" on in Gaming 3D Application Settings in CCC. Also check the "refresh rate" is correct for your monitor in Desktop Properties in CCC.

Having said that, my monitor is running at 60Hz (default) 1920x1200 and I have Vertical refresh off unless application specifies. Triple buffering was also off on my cards. Is it possible there are 'in game' advanced video settings that are affecting your screen?
m
0
l
January 14, 2013 11:54:06 PM

RatsaK said:
According to other posts on screen tearing: If you are still getting screen tearing, you might look at turning "wait for Vertical refresh" and "Triple buffering" on in Gaming 3D Application Settings in CCC. Also check the "refresh rate" is correct for your monitor in Desktop Properties in CCC.

Having said that, my monitor is running at 60Hz (default) 1920x1200 and I have Vertical refresh off unless application specifies. Triple buffering was also off on my cards. Is it possible there are 'in game' advanced video settings that are affecting your screen?


Turning VSync on helps tremendously in my case for screen tearing. It seems to be a highly debated topic though. Some people get really bad lag when they turn it on apparently. [rig listed in earlier post]
m
0
l
January 17, 2013 12:32:19 AM

Ok guys, here's the update. I finished the installation of this aftermarket vga cooler this morning for my hd 5870:

http://www.newegg.com/Product/Product.aspx?Item=N82E168...

While it brought my temperatures down a WHOLE lot, I STILL experienced the crashes when I tested skyrim under load shortly after. So I am now inclined to agree with you ratsak. I always knew my core temperatures were fine, but I thought that some of the other chips on the card could be overheating and that an aftermarket cooler would solve things. Alas, it was not the case.

So I decided to give ratsak's underclocking solution another try...and everything appears to be running ok so far, crash-free! (I disconnected the hdtv to bring my setup to single monitor and that seemed to resolve the issues I had earlier with screen tearing and whatnot--though I will try fiddling with the settings that were suggested earlier if I get problems again when I plug it back in).

Now I'm thinking that maybe the root cause of our crashing problems could be with the power supply? Check out the following thread:

http://www.tomshardware.com/forum/323969-28-computer-cr...

This guy's rig crashed when his system was under load, and it turned out that the cause was the power supply. I am reluctant to spend even more money than I already have chasing this problem, so for now I'm just going to keep my system underclocked. Ultimately, however, this may be the direction we need to look if we want to get at the core of this problem.
m
0
l
January 17, 2013 1:25:01 AM

It is always good to keep an open mind. In my case (literally) the power supply is more than adequate - Corsair 1000W. Keep us updated mrxenyo and good luck.
m
0
l
January 20, 2013 8:24:56 PM

Just a quick update: it's now been several days of gaming under full load in Skyrim, and I haven't had a crash at the underclocked settings. (750/1100). Thanks again for the fix ratsak.
m
0
l
February 2, 2013 2:01:03 AM

EDIT: I thought I found a solution, but didn't actually as the crashes recurred. Ultimately, I'm not sure if the better power cables helped or not, but I'll leave my original post alone just to document what I attempted anyway.

Ok, major update. Turns out it was my power supply cables the whole time. I feel rather foolish to admit this, but I lost a few of the modular pci-e cables that came with my power supply after I moved a while back (OCZ ModXStream Pro 600W Modular Power Supply) and as a result had to rely on a molex-to-pci converter cable like this one http://www.computer-answers.ca/wp-content/uploads/2012/... for the past 6 months or so. Since I used it successfully with my last graphics card (GTS 250) I figured it'd be ok at the time, but the hd 5870 is way hungrier as it turns out and evidently requires the direct connection to my psu for both the 6-pin connectors.

I didn't think of the power supply right away when I was troubleshooting this issue, since my psu is pretty decent and is a known good quality psu even for overclocking (though I'm not overclocking). But after reading so many posts of people with this problem having power supply issues, I thought about it more and more and then the obvious deficiency of my cables occurred to me at long last.

OCZ is AMAZINGLY accommodating when it comes to replacing modular power supply cables, by the way. I was fully prepared to buy a new set on ebay, but after turning up some good stories in forum searches, I figured what the hell and wrote their RMA people a note explaining my situation and offering to buy the cables if necessary. Well Allen Chung of OCZ wrote me back and actually mailed me two new modular pci-e cables for the psu for FREE. For the cables AND the shipping. (Sweet!!!)

Anyway, I deliberately refrained from posting here right away as I have had several "false positives" in the past and wanted to make absolutely certain that I had accurately diagnosed the problem, and would not experience a crash again after posting. I have not. (It's been about 2 weeks, gaming under full load every day or two. No gray screens or crashes of any kind).

What my experience suggests to me is that many of you experiencing the gray screen crash probably have power supply problems in your cases as well...

Anyway, hope to hear about your situations and I hope this helps some of you guys. Let me know if you have any other questions about how I resolved the issue and I'll be glad to help however I can.

(Minor edits for grammar and to add an additional paragraph for clarity)
m
0
l
February 11, 2013 2:21:59 AM

I am disappointed to report that I unfortunately experienced another crash. I'm back to operating at the underclocked 750/1100 configuration that has been stable for many of us. At this point I have given up on experimenting and just want to enjoy this card and gaming, so I'm going to be running it at 750/1100 for gaming and switch to 850/1200 for movie-watching.

I hope you guys have better luck learning more about this problem. I'll be following this thread in case anybody else has new or better information. Thanks again for everything.
m
0
l
February 11, 2013 7:36:49 AM

Sorry to hear your news. At least dropping the clocks back a bit should give you a stable card until there is a better fix.
m
0
l
February 27, 2013 6:00:02 AM

Sorry for the thred necro but i had this problem myself but fixed it buy removing my dual-boot (W7/W8) if you have this i fear it might be a driver clash because when I took it off I don't get anymore crashes and i took my Gigabyte HD7850 Windforcex2 OC up to 1050/1300 clocks and it works great no rcashes anymore but I also ran my GFX cards fans @100% for a few hours to make sure that the fans were working
m
0
l
March 25, 2013 11:45:36 AM

Thomzey said:
Sorry for the thred necro but i had this problem myself but fixed it buy removing my dual-boot (W7/W8) if you have this i fear it might be a driver clash because when I took it off I don't get anymore crashes and i took my Gigabyte HD7850 Windforcex2 OC up to 1050/1300 clocks and it works great no rcashes anymore but I also ran my GFX cards fans @100% for a few hours to make sure that the fans were working


Oh wow, really? I have a dual boot setup as well. To be more precise, it was for a HyperSLI setup I ran for my last video cards (pair of GTS 250s) that I replaced with the Radeon. I'll try completely dismantling all of the HyperSLI boot stuff and report back.

Thanks for the suggestion!
m
0
l
September 2, 2014 3:47:51 AM

Hi mrxenyo

I have bought a R9280x Toxic and I am also experiencing this grey screen issue.
So far, I only experienced the issue while playing Dota 2 for about 3 - 4 hours.
Have you manage to find out what this could be related to yet?
m
0
l
October 7, 2014 2:22:48 PM

Sadly, I have the grey issue too...
PC spec:
ASUS P7P55 LX
Intel Core i7 860
4x2GB Corsair Dominator
R9 270x Vapor-X
Chieftec A135 series - APS-650C - 650W

Some times, the grey screen appears 40-50minutes after started playing LoL. Some times 2-3 hours...
I have read the comments, and will try to underclock it, but as I read, it doesnt help too much.. I will try replacing the 6 pin cables too, and see if that helps, but I doupt... Wish me luck.. If you find any other solutions, please reply, because it's getting annoying. :( 

Edit: The video card is PCI-E 3.0, but the motherboard is PCI-E 2.0. Can that be an issue ?
m
0
l
!