Witcher 3 Crashes and artifiacts GTX 1080

melenie

Commendable
Jun 9, 2016
90
0
1,660
I'm experiencing some rather strange behavior whilst trying to play the witcher 3 on my new PC.

There are random black artifacts and screen tears which are appearing. The game crashes every 10mins or so.

Things I have tried (in order)

.Removed in game mods
.Set GPU to run back at stock
.Set cpu clock back to default
.checked gpu drivers (up to date ver 368.81)


Any other suggestions guys?


Link to my entire system specs
http://

UPDATE:

Just some extra info.
No I haven't tried any other games. This is the only game I currently have installed. I've also done a google search and haven't seen any other posting any issues with this current driver and the witcher 3.

Also worth mentioning. I am using a pcie extender with my gpu. Could this be the my issue? It came out of box fairly kinked.


13835982_2086105798281583_684723227_o.jpg


20160718_170030.jpg
 

melenie

Commendable
Jun 9, 2016
90
0
1,660


It's a complete new system. However..in saying that, I am using the pre existing SSD/HDD. All I did was plug it into the new motherboard etc. Could this be causing my issues? Please also refer to the edited info up top in regards to pcie extender :)
 

Geekwad

Admirable


Ah, that would explain it. I would very highly recommend doing a clean install of the OS on the SSD.
 

melenie

Commendable
Jun 9, 2016
90
0
1,660


it never even crossed my mind. I got lazy, plugged everything in after a long grind of getting this baby ready! Those extravagant bends really tested me, especially my piece from res to pump haha.

I'll go ahead now and try the fresh install. I'm assuming I'm going to have to purchase another windows key?
I'll also need to make a windows disk as i don't have one :/. The old pc was upgraded from windows 7. I used a windows 7 disk image (no idea where it's gone) . I then purchased a legit key online.

Does Microsoft have somewhere I can make a windows 10 disk image? is the free windows 7 to 10 upgrade still running? If so, I would assume it would be cheaper for me to purchase another windows 7 key and then free upgrade to 10?
 

Geekwad

Admirable
You can get the media creation tool here:

https://www.microsoft.com/en-us/software-download/windows10/

And you should be able to use it activate Win10 directly if it was a full version key. No need to install 7 to upgrade to 10 (and in fact is best if you don't).
 

melenie

Commendable
Jun 9, 2016
90
0
1,660
Okay so I've done a fresh install. Windows 10 is running. Formatted my SSD where the OS is installed. Did a disk defrag of my HDD and also a optimization of my other SSD.
I no longer get in game crashes. I've been playing 1hr without a single crash. However I still get weird things appearing on the screen.
Random flashes of red dots. Kinda looks like a red LED is in my screen and it randomly flashes and dissapears etc.

I'm really starting to think it's this pcie extender -.-
 
is this led flashing on the same spot every time? check your monitor and connection to the montior. How did you connect it? VGA, DVI, HDMI, any adapter between?

try it out of the case without pcie extender

bios up to date?

check temps with furmark burn in
 

melenie

Commendable
Jun 9, 2016
90
0
1,660

There a few places which flash which aren't always in the exact same spots but seem to be in the same areas (top middle and bottem left).

Temps are more than fine. I run at 34c heaven benchmark overclocked.

I have quite a problem testing with the card plugged into the mobo and not where it currently is. I have a full custom watercooling loop with acrylic pipes. I'd need to drain my loop and create some new tubing etc :pfff:

Such a pain in the ass. I think I'm just going to go ahead and purchase a new pcie extender and go from there. If the problem persists, I'll need to drain the loop for a 2nd time, take the block off the card, run the air cooling setup and test it on a different computer to see if the card itself is faulty?

:/
 

Geekwad

Admirable
When you did the new OS install on SSD, did you update all the other hardware drivers.....chipset drivers?

Would also be helpful to know your complete system specs, including model numbers. The link you provided (to CPU-z) does not work.

 

melenie

Commendable
Jun 9, 2016
90
0
1,660
Been away on holidays. Got back last week.
Purchased a brand new pci-e extension cable as I did some searching around and found people having issues with the included thermaltake cable with the core p5 cases.

Of course this meant have to drain my entire loop, rip apart all the blocks and clean everything. Purchase new thermal paste and new liquid etc. This was another hit to my wallet as well as many hours of work.

Still the same issues.

All drivers are up to date. This also includes any chip-set drivers etc. I even went out and purchased a new hdmi cable just in case the other was faulty.

I decided I would try something different. Opened msi afterburner, underclocked my card dramitaclly. -400 core clock and -500 mem clock.

I've lost about 40fps in game but no red flashes, no tears no issues.

Conclusion? Faulty card?
Able to RMA a card which has had stock cooler removed and water block attached?

FML.
 
bios of the motherboard is up to date now?

you can try to rma it, but it could be that you damaged a chip while installing the waterblock on the gpu.

does the water block cool the vram too?
couldn´t open your cpu-z link, could you post your specs please?
especially the water cooling solution you installed
 

melenie

Commendable
Jun 9, 2016
90
0
1,660
Everything is up to date. All drivers, windows updates, gpu. Mobo was flashed with a bios update which seemed to be irrelevant but i flashed it nevertheless.

In regards to any physically damage being dealt to the vram during the installation process of the water block, I do believe this is highly unlikely. Having built more pc's then I can count on two hands, I would say I'm definitely accustomed to taking extra care when working around exposed pcb's.

Gpu waterblock covers vram.

Windows 10
Asus Maximus VIII HERO
Intel i7 6700k
MSI GTX 1080 Gaming X
G.SKILL TridentZ Series (2 x 8GB) DDR4 3600
EVGA Supernova G2 750W 80+

I have a full custom watercooling loop;

Raditator: EK-Coolstream PE 480 Quad
Resevoir: EK-RES X3 250
Pump: EK-XTOP Revo D5
Fans: Push pull configuration on radiator. 4x Thermaltake RGB RIING 120mm fans in push
4 x Corsair SP 120 fans in pull

CPU Block: EK supremacy evo
GPU Block: EK-FC1080 GTX TF6
Ram Block: EK-RAM Monarch X4

Thermal Grizzly Hydronaut grease used on all blocks.

All EK black nickel fittings with EK 10/12 Acrylic tubing

Using mayhems aurora 2 silver with mayhems emerald green dye.

No expense has been spared on my cooling solution.

Is there any specific application which can test modern VRAM effectively? I came across "video memory stress test v1.7.116" but it seems to be outdated, only testing 1gb vram. Also, the instructions aren't very clear etc.

I've gone ahead and set the card clocks back to default. Problem persists. I then set the mem clock to -50 and artifacts are removed.

photo sharing
 
BGAs can be very touchy. Physical damage to the soldering balls can be possible, while removing the MSI cooling solution.

Did you wear a grounding strap?

You could try to raise the power limit a bit higher and set the clock back to default.

this old occt test could be interesting:
https://www.raymond.cc/blog/having-problems-with-video-card-stress-test-its-memory/
 

melenie

Commendable
Jun 9, 2016
90
0
1,660
Grounding wasn't worn during the installation of the block.

I ran evga oc scanner with the artifact scanner and after 30mins 0 artefacts were detected.

I've been able to play witcher atm with no artifacts with the mem clock slightly underclocked which doesn't seem to have too much of an impact on performance.

New issues I'm experiencing now though. Some sort of weird screen tear / blurry line which jitters around about 1/3 of the way up the screen (horrizontally) and then dissapears.

Also, the one which is annoying me the most. When I die in game and have to reload or I pause to menu and then resume, I then enter back into the game with ridiculous amounts of lag and performance issue. Frame drops to 20fps and massive amounts of lag. Decreasing in game settings to low does not fix the issue. Only way to rectify this occurrence is to completely exit the game and then start over. Any suggestions? This is all driving me insane.
 

melenie

Commendable
Jun 9, 2016
90
0
1,660
Have switched from hdmi cable to display port. Also, turned on V-sync and screen tearing has been fixed. Now I'm wondering if having V-sync off could be the reason as to the red dot artifacts? Didn't even realize it was off.

Still not sure what is going on with the massive lag and frame drop after loading screen or game pasued and then resumed. Happens randomly. I did however notice that when it happens and I then change some video settings and then revert back, it fixes it.

hmmm


 
run 3dmark basic edition (firestrike default settings) and post the link which comes up at the end in your browser

furmark stress test made yet?

reinstall nvidia drivers by uninstalling with DDU uninstaller and install newest drivers from nvida.com

if you are resuming the game, how high is the frequency of the CPU and the graphics card?

sreen tear looks like this? http://www.144hzmonitors.com/wp-content/uploads/2015/04/Screen-tearing.jpg
v-sync -> on
 

melenie

Commendable
Jun 9, 2016
90
0
1,660
Last week my system died. PC turned on but motherboard wouldn't post. After painfaully draining my loop and dismantling the pc, I did a basic breadboard with the mobo with basic components to verify what was causing the issue. The motherboard was dead. Seemed to be bricked. RMA, got a replacement board in store, reassembled the system.

I no longer have any issues with screen tears/ red particles. Interesting to see that the issue might have been the motherboard. So odd how the system ran fine (except for the ingame artifacts) for weeks and then one morning, Just failed to post.