Out of Solutions

[Shade-0117]

Hey everyone, I do not post on forums often but I find myself in need of help. 

 

Normally I am the guy that everyone goes to for PC help, but I find myself at a loss for the problems that I am having with my own system.

 

What keeps happening is that the AMD drivers keep crashing. Either they crash, pop up the message that the drivers have crashed and recovered and then the game that I was playing crashes in turn (duh) or the whole system goes down; screens go black and the only solution is to hard reset. 

 

I'm sorry I cannot be more specific than this but it is happening in every game that I play. Recently I have been replaying alot of games so here is the list of games that I have been having issues in: Mass Effect, Mass Effect 2, Mass Effect 3, Dragon Age: Origins (And all DLC/Expansions as they are the same game), Dragon Age II, Dragon Age: Inquisition, Grand Theft Auto V, Company of Heroes 2, World of Warships, Fallout 3, Fallout: New Vegas, Homeworld Remastered, Hitman: Absolution, Borderlands: The Presequel... yeah. I am pretty sure that the problem is down to something with the drivers or my system. 

 

My system setup is kinda complicated. I recently changed over to using my secondary tower as the main had me throwing fits of rage. But the secondary is giving me the same problems, if less frequently. And on top of that I have changed the secondary a bit to compete with the power my primary had. (Note: All 3 of these setups have been giving me the SAME problems)

 

Original Setup:

Gigabyte G1.Sniper A88X

AMD Athlon II X4 860K

Corsair LP 2133 RAM (2x 4GB)

PowerColor Radeon R9 290X

PowerColor Radeon R9 290 

CoolerMaster Hybrid Pro 850W PSU

 

Secondary Setup:

MSi FM2-A85XA-G65

AMD A10 6800K

Radeon RP1866 RAM (2x 4GB)

HIS Radeon HD7750

Fractal Design Tesla 650W PSU

 

Current Setup:

MSi FM2-A85XA-G65

AMD A10 6800K

Corsair LP 2133 (2x 4GB) and Radeon RP1866 (2x 4GB) RAM (I know I am running 2 different kits, see later explanation)

PowerColor Radeon R9 290

Fractal Design Tesla 650W PSU

 

Now, to this point I have been dealing with the issue but it is really staring to grind my gears -- namely cause I have started playing MP games with my friends again. So, what have I tried? Here is another list, and this is just what I can remember having tried:

 

1. Uninstalling and reinstalling the AMD drivers, both the beta and non-beta versions (I use DDU to completely remove the drivers)

2. Underclocking my GPU

3. Underclocking my CPU

4. Changing RAM

5. Removing the Crossfire setup

6. Running MSi Afterburner to monitor temps

7. Uninstalling and reinstalling DX

8. Updated BIOS on motherboard (Both on the previous and current setups)

9. Updated BIOS on the R9 290

10. Changing out 8 pin connectors from PSU to GPU

11. Flipping switch on the 290 from normal mode to quiet mode and vice versa

12. Various fixes that are specific game related (Disabling in-game for programs, overlays, specific profiles, blah blah)

 

So, for the RAM configuration that I am currently running, I found that I am having the same issues regardless of which RAM kit I have in my system and the fact that I am having the same issue with BOTH kits installed, directs me to thinking that the RAM is not the issue. I do know that RAM problems can emulate problems with other parts of your system... but in this case I think that I have already proved that it is not the RAM. 

 

Also, you will note that I have not mentioned what OS I am running, that is because I am currently running Windows 10. But, I have been having the same issues since I was running Windows 7. So I have had these issues with Windows 7, 8, 8.1, 10 Technical Preview, and 10 Official Release (All fresh installs BTW). Like the RAM, this leads me to believe that my problem is independent of the operating system that I am running. (Does kinda tempt me to try it with Linux though)

 

I think this is about all that I have to say about it. If you have any ideas, please post them cause I am at the end of what I know what to do. 

[PlayStation 2]

That's odd.

Drivers on my laptop have never really crashed on me.

 

Maybe it's... the hard drives?

[pokechat8978]

-snip-

 

Try Catalyst 15.8 drivers.

http://support.amd.com/en-us/kb-articles/Pages/latest-catalyst-windows-beta.aspx

[Shade-0117]

I am currently running the 15.8 drivers. I reinstalled earlier today :-(

[BlueChinchillaEatingDorito]

Does this only occur during gaming? How about under normal use like browsing the web, editing documents, etc.

[Deagon]

Set the GPU and CPU clocks to their base settings, don't under clock them, they hate doing that. 

 

Uninstall the drivers before you try reinstalling them sometimes little bits are left over. 

[blackadder]

I am currently running the 15.8 drivers. I reinstalled earlier today :-(

checked hard drives for bad sectors, maybe a conflict with a program with catalyst.

Worth a shot to try this, open admin cmd prompt and type the following

Sfc /scannow

^should have 0 errors.

[Shade-0117]

It seems to only happen during gaming yes. I have not noticed it during any other usage, but this system is generally only used for gaming. I watch YouTube, check email, and such on my laptop. Which I guess I should add that I do not have these problems with.

 

My laptop is an Origin EON-13S.

Intel Haswell Core i5 4340M

nVidia GeForce GTX 860M

8GB DDR3 1600MHz RAM

 

Please no Intel/nVidia vs. AMD flame wars... I want my computer to work, not have a 200+ post long thread on what people think is better. 

[Shade-0117]

Set the GPU and CPU clocks to their base settings, don't under clock them, they hate doing that. 

 

Uninstall the drivers before you try reinstalling them sometimes little bits are left over. 

 

I am running at stock clocks now, the underclocking had no effect so I reverted. Also, the utility I use, Display Driver Uninstaller (I think it is by TechPowerUp), uninstalls the drivers for you then restarts with Windows set to not automatically find and download whatever it thinks it the best drivers for your display. 

 

checked hard drives for bad sectors, maybe a conflict with a program with catalyst.

Worth a shot to try this, open admin cmd prompt and type the following

Sfc /scannow

^should have 0 errors.

 

I will try that and report back. 

[GhostUN7]

Set the GPU and CPU clocks to their base settings, don't under clock them, they hate doing that. 

 

Uninstall the drivers before you try reinstalling them sometimes little bits are left over. 

With dealing with this on TS, Shade has done this, and sadly it didn't work...

 

 

I see I was beaten to the response... Anywai...

 

 

If i were to guess in trouble shooting terms in a list I would go by...

 

Power Supply: But hes using a new system

Memory: but he had this problem with different memory

HDD: Had this problem with different Hard Drives

OS: Had this problem from Windows 7, and fresh install up to 10, problem still occurs

Mother Board: Different system, not the same motherboard

 

The only thing that is left in my view is Software, but I doubt it is drivers... Maybe something else conflicting against AMD drivers?

[Deagon]

With dealing with this on TS, Shade has done this, and sadly it didn't work...

 

 

I see I was beaten to the response... Anywai...

 

 

If i were to guess in trouble shooting terms in a list I would go by...

 

Power Supply: But hes using a new system

Memory: but he had this problem with different memory

HDD: Had this problem with different Hard Drives

OS: Had this problem from Windows 7, and fresh install up to 10, problem still occurs

Mother Board: Different system, not the same motherboard

 

The only thing that is left in my view is Software, but I doubt it is drivers... Maybe something else conflicting against AMD drivers?

I used to have a weird conflict with my board it had a GPU made by a different company than the GPU add in card, i had to unstiall all the drivers for the boards GPU to even get the Add in card to work. Perhaps something like this? It does seem strange. might be that the cards aren't supporting the games, or something isn't compatible, but it seems strange that it happens across 3 systems and configs. 

[Shade-0117]

checked hard drives for bad sectors, maybe a conflict with a program with catalyst.

Worth a shot to try this, open admin cmd prompt and type the following

Sfc /scannow

^should have 0 errors.

 

"C:\WINDOWS\system32>sfc /scannow

 

Beginning system scan.  This process will take some time.

 

Beginning verification phase of system scan.

Verification 100% complete.

 

Windows Resource Protection did not find any integrity violations."

 

Well it was worth a shot

 

 

I used to have a weird conflict with my board it had a GPU made by a different company than the GPU add in card, i had to unstiall all the drivers for the boards GPU to even get the Add in card to work. Perhaps something like this? It does seem strange. might be that the cards aren't supporting the games, or something isn't compatible, but it seems strange that it happens across 3 systems and configs. 

 

My original config with the Athlon II X4 860K (FM2+ Socket) does not have an iGPU. The CPU is a standard A10 but with the iGPU disabled in manufacturing. The A10 6800K (FM2 Socket) that I am currently running has an AMD HD 8670D (Confirmed with CPU-Z) iGPU, but I am running my second monitor with that right now and it never has a fault. When the GPU drivers crash, that screen stays up. That tells me that the driver for the R9 290 is the one that is crashing (They are separate drivers with separate HDMI audio driver lanes and everything in my device manager). 

[Deagon]

 

 

"C:\WINDOWS\system32>sfc /scannow

 

Beginning system scan.  This process will take some time.

 

Beginning verification phase of system scan.

Verification 100% complete.

 

Windows Resource Protection did not find any integrity violations."

 

Well it was worth a shot

 

 

 

My original config with the Athlon II X4 860K (FM2+ Socket) does not have an iGPU. The CPU is a standard A10 but with the iGPU disabled in manufacturing. The A10 6800K (FM2 Socket) that I am currently running has an AMD HD 8670D (Confirmed with CPU-Z) iGPU, but I am running my second monitor with that right now and it never has a fault. When the GPU drivers crash, that screen stays up. That tells me that the driver for the R9 290 is the one that is crashing (They are separate drivers with separate HDMI audio driver lanes and everything in my device manager). 

 

Um, have you tried doing a driver "rollback" ? if you are lucky its just the current drivers are conflicting with something and an older driver could be better. 

 

conceitedly the issue i had was with an MSI board with an fm2 socket and a radeon 6850 

[Shade-0117]

Um, have you tried doing a driver "rollback" ? if you are lucky its just the current drivers are conflicting with something and an older driver could be better. 

 

conceitedly the issue i had was with an MSI board with an fm2 socket and a radeon 6850 

 

I have not tried, but I have been having this issue since before the AMD Catalyst drivers called themselves "Omega"... so to rollback to a point that I have no knowledge on would be... pre 14.xx? 2 year old drivers? I'd prefer to find another solution.

 

I don't want to sound like I am dismissing your idea, but I have thought of this already and... yeah 2 year old drivers? I'd rather not. I may try it if something else doesn't pop up and fix my problem but I consider that a last resort, barring someone doesn't give me a really damn good reason to do so.

[Deagon]

I have not tried, but I have been having this issue since before the AMD Catalyst drivers called themselves "Omega"... so to rollback to a point that I have no knowledge on would be... pre 14.xx? 2 year old drivers? I'd prefer to find another solution.

 

I don't want to sound like I am dismissing your idea, but I have thought of this already and... yeah 2 year old drivers? I'd rather not. I may try it if something else doesn't pop up and fix my problem but I consider that a last resort, barring someone doesn't give me a really damn good reason to do so.

Two years, yeah i'd have to agree there. 

[Universal_Cereal_Bus]

There is an old ati article http://support.amd.com/en-us/kb-articles/Pages/737-27116RadeonSeries-ATIKMDAGhasstoppedrespondingerrormessages.aspx

 

But it sounds like the problem could be the environment as well- is it abnormally warm or dusty or humid? Does the crash occur immediately after starting a game or after minutes of play- and which driver is shown to crash?

 

The other constant could be DirectX it is going to be the common factor among all of these games. Each game tries to isntall its own distribution of directx and various other packages- do you think its possible that installing ALL of the games together may have caused a problem?

 

Are you running the games from another drive- as in same installs? Perhaps try a computer with only ONE of the games installed (if you can, fresh install windows and go game by game).

 

also make sure that all of your pci contacts are clean- if one of the original cards had any residue on them and you swapped it into the other board, you could have covered some of the pci pins to where full bandwidth is not possible for any of the cards. Otherwise, check peripherals? Some monitors react strangely when going full screen.

 

on that note- try running a game in windowed mode

[Shade-0117]

There is an old ati article http://support.amd.com/en-us/kb-articles/Pages/737-27116RadeonSeries-ATIKMDAGhasstoppedrespondingerrormessages.aspx

 

But it sounds like the problem could be the environment as well- is it abnormally warm or dusty or humid? Does the crash occur immediately after starting a game or after minutes of play- and which driver is shown to crash?

 

The other constant could be DirectX it is going to be the common factor among all of these games. Each game tries to isntall its own distribution of directx and various other packages- do you think its possible that installing ALL of the games together may have caused a problem?

 

Are you running the games from another drive- as in same installs? Perhaps try a computer with only ONE of the games installed (if you can, fresh install windows and go game by game).

 

also make sure that all of your pci contacts are clean- if one of the original cards had any residue on them and you swapped it into the other board, you could have covered some of the pci pins to where full bandwidth is not possible for any of the cards. Otherwise, check peripherals? Some monitors react strangely when going full screen.

 

on that note- try running a game in windowed mode

 

I read through that ATI article and unfortunately did not see anything that I think is pertinent.

 

As to DirectX, I doubt it. I have tried reinstalling it many times over the past week and nothing seems to fix the problem.

 

In the first setup that I posted, I had games installed over 3 different hard drives (OCZ Agility 4 240GB SSD, Western Digital Black 1TB HDD, and a Western Digital Green 2TB HDD). They all had the same problem. In my current setup I have all of them installed to a Western Digital Black 1TB, but I can try installing a few games on my boot drive, OCZ Agility-3 120GB, and see if that makes a difference.

 

I will check the contacts on the motherboard next, that is a good suggestion and I have not done that. I do look at the pins on the cards after I pull them out but I did not think to look at the slot itself.

 

Windowed mode is a no-go. Another of the things that I have tried. I get the same issues with Fullscreen, Windowed Fullscreen, and Windowed modes.

 

I'll report back after I check the PCIe slot pins on my mobo. Gonna check both my boards in case it has happened with either or both of them. GhostUnit also found an article online that I had not seen yet, http://www.tomshardware.com/answers/id-2033153/explain-works-290-black-screen-fix.html. If you scroll down to the 7th or 8th post (The really long one) I am following that right now. I do usually run my GPU's with the power limits raised, but I do not mess with the core voltages like his fix/hypothesis/research/whatever suggests. With those fixes in place I have been running MSi Kombustor on a GPU core burn test for 45 minutes with no crashing and a 10 minute run on the GPU memory with no crashes. I'm moving to 3DMark to run all those tests to see if I get any crashes. If that is successful I guess I'll try gaming.

[Shade-0117]

No go. No crashes in Kombustor or in 3DMark. But I started up Grand Theft Auto V and started messing around in that. After about 10 minutes it crashed. Black screen, Windows made the noise of a notification popping up (Like it always does for the display drivers crashing and 'recovering'), then it made the noise a second time, and the screen never came back up. That tells me the drivers completely failed and could not reinitialize.

 

PCIe slots were clean as well, no gunk or anything out of the ordinary that I could see.

 

I'm seriously at a loss now. No idea what is going on. I don't think it is a hardware fault; temps are in acceptable levels even for that last GTAV crash I was at 71C, so many different iterations of hardware, underclocking, overclocking, increasing voltage limits, reflashing BIOS...

 

Thinking about selling all this hardware and building a whole new system.

[Shade-0117]

It seems that no one else has any input onto my issue, but I will continue to post (Specifically if I resolve the problem) so that if someone else is having the same issue, they might find something useful in my trail.

 

I've found an article on the Microsoft support page that I am now investigating.

https://support.microsoft.com/en-us/kb/2665946

[Shade-0117]

Okay, I seem to have fixed the problem, not optimal, but at least I can play games for the first time in a week.

 

So, since my last post I have fixed the "Display Driver has Crashed and Recovered", but in fixing it there was another issue that popped up. In order to fix the first problem of the drivers crashing with no error messages other than the driver has crashed and recovered, I followed https://support.microsoft.com/en-us/kb/2665946.

 

The problem appears to be that the graphics card hangs just long enough before outputting a new instruction to the monitor that Windows deems that it has crashed. Windows then closes the drivers and restarts them in an attempt to get the graphics card to continue working. Problem is that the graphics card is about to output a new instruction and Windows has no need to intervene, just wait a little bit longer. Windows restarting the drivers is what causes the error messages and the drivers being restarted is what crashes the game. What this fix does is just tells Windows to wait longer for the graphics card to output an instruction. I used the 8 value that was recommended in the article and it appears to be working, though I am sure that you could tweak that number if you still have issues.

 

Using that, I tested with GTA V again and had a new issue. It started crashing as it had before, but now instead of the "Display Driver has Crashed and Successfully Recovered", I was getting a message saying "[Application] has been blocked form accessing the Graphics Hardware" AND a pop up message behind the game that stated, "ERR_GFX_D3D_INIT". Lol what? Okay...

 

After research I found that this is a common issue with GTA V, so I decided to use a new game as my test game. Dragon Age: Inquisition started giving me the same issue. Namely, I was talking to the Enchanter in the outcropping just behind Redcliffe Village in the Hinterlands. But it was not just talking to her, after getting past that part I was still having the issue every 5-10 minutes.

 

I found: https://support.rockstargames.com/hc/communities/public/questions/203458477-ERR-GFX-D3D-INIT-Crash-BSOD

People were reporting, "This worked for me: I am running the game without VSYNC and in Borderless window with tesselation OFF. I couldn't run the game for more than 5 mins with these settings on before I disabled them". This worked for me, but only to fix the "ERR_GFX_D3D_INIT" pop up. DA:I and GTA V would now go longer ~30 minutes before crashing, and would no longer output the "ERR_GFX_D3D_INIT" box, but they still had the "[Application] has been blocked form accessing the Graphics Hardware"

 

I found this: http://steamcommunity.com/app/200110/discussions/0/35219681738161861/#p3

In here they say something about it possibly being PhysX (Which I am running an AMD card so that is a reasonable assumption) but neither GTA nor DA:I have the ability to disable PhysX effects in the game settings. The other thing that was mentioned is that someone underclocked their GPU by 100 MHz from the Core and Memory clocks. I was still running the stock clocks with the power limit and voltage changes that I mentioned in a post before. I decided to try the 100 MHz underclock on Memory and Core with those same Power and Voltage changes from before.

 

Boom. 45 minutes no crashing in DA:I, but my performance was suffering something awful. I decided to take only 5% from my Core and Memory clocks, which minimized the amount of performance decrease I was seeing, but still results in no crashing. I played GTA V online with some friends for about 2 and a half hours last night, and then went into Hitman :Absolution for a mission (Which I could not previously get to the next checkpoint without crashing), and then went into Dragon Age: Inquisition for 3 hours. No crashes, no hiccups, nothing to indicate system instability.

 

It looks like I have fixed it, and hopefully if someone else is reading this because of the same issues you can follow my breadcrumbs.

 

Here is exactly what fixed it for me:

1. Do the registry changes found on this article https://support.micr...n-us/kb/2665946

2. Download and install MSi Afterburner. Go to settings and tell it to run on Windows startup (Sometimes it won't work and you have to launch it manually). Also check "Unlock Core Voltage Control" and "Force Constant Voltage". Then using the sliders put your Core Voltage at +25 mV, and power limit at +5%. Then take between 5 and 10% from your Core and Memory clocks and set those sliders to the resulting numbers (If you click on the numbers you can manually input numbers as opposed to fighting with the sliders. If you still have the crashing issues, try taking a little more off the clocks.) I'll post my resulting numbers below. It might be a good idea to manually set a fan curve too, just to make sure your GPU stays cool, I personally do not let it get about 75C and I target 70C as a max operating temperature.

3. Disable Tessellation, VSync, and run games in Windowed Borderless (I'm not sure if you really have to do this step, but normally I run in Windowed Borderless anyways, I almost always have Vsync off by default, and the only thing that I had to change was Tessellation - of which did not change my performance or visual quality by any amount that I could notice.)

 

Here are the numbers that I am running my R9 290 at now:

Core Voltage: +25 mV

Power Limit: +5%​
Core: 925 MHz

Memory: 1185 MHz

​

the stock number for the chip are:

No boost to Core Voltage or Power Limit

Core: 975 MHz

Memory: 1250 MHz

 

I really hope this will help someone else save the time it takes to fix this crap.

[Shade-0117]

And nope. After starting my PC this morning, I started having the "[Application] has be clocked from accessing the Graphics Hardware" error again. Apparently something changes between last night and this morning. 

 

I'm done, both systems are getting torn apart and sold.

[Deagon]

And nope. After starting my PC this morning, I started having the "[Application] has be clocked from accessing the Graphics Hardware" error again. Apparently something changes between last night and this morning. 

 

I'm done, both systems are getting torn apart and sold.

Did you try a system restore? might help. Otherwise, a clean install might help. 

[Elite14]

And nope. After starting my PC this morning, I started having the "[Application] has be clocked from accessing the Graphics Hardware" error again. Apparently something changes between last night and this morning. 

 

I'm done, both systems are getting torn apart and sold.

I have been having this same problem with my brothers pc for the last year or so and it's extremely annoying and I have tryed everything under the sun, I will try this and let you know if it works