PC freezes with grey screen with vertical lines during Gaming

Status
Not open for further replies.

omarmoustafa280x

Reputable
Apr 6, 2018
12
1
4,515
So over the last few days I have experienced constant crashes & freezing during playing any game or benchmark & the same problem happens for each app..PC freezes with grey screen with vertical lines & buzzed sound, sometimes it's red or yellow but with the same vertical black lines & few times it freezes with black screen but no blue screen error appears..this problem happens after only 2 or 3 minutes of running the game or benchmark & I have to make a hard reset in order to get things back to normal but also artifacts appear at bios start up & some vertical lines at windows logo then a black screen & I can't get to desktop..after unplugging the power cord & vga cable or after few resets pc boots normally without artifacts & I can get to desktop...So I have made many tries in order to get thing back to normal but in vain & I'll state them after the rig specs & status.

The specs are:

CPU: i5 4590
GPU: Sapphire R9 280x Tri-X: 1020 core clock - 1500 memory clock
PSU: EVGA 600W Bronze
Motherboard: Gigabyte B85-HD3
RAM: 2X4 Kingston hyper x fury
HDD: 1 TB WD blue

Temperatures:

CPU: avg 37~40 at idle .... avg 60 at full load

GPU: avg 39~46 at idle .... avg 70~80 at full load

Voltages:

GPU: 0.850 at idle ... 1.256 at full load for core clock & 1.600 for memory clock

So I have made the following failed attempts:

- reinstalled gpu driver many times with different versions.
- reinstalled windows 10 two times with different builds.
- installed windows 7 & then windows 8.1 which is the one am working with now.
- removed cmos battery & updated bios to latest version.
- underclocked gpu to 940 for core clock & 1250 for memory clock without changing the voltage.
- raised fan speed to 70% in full load instead of 53%
- made a memory test with memtest86+ but for one pass only for each stick & there were no errors at all.
- checked hdd with chkdsk & there were no errors.
- made stress test for cpu with occt & prime & there were no errors & temps were normal as I stated 60C
- & this test was an interesting one as I made a stress test for video memory with "video memory stress tool" & there were no errors at all..I was surprised as those artifacts indicates that my gpu is going bad & there is something wrong with vram.
- tried another monitor with hdmi as I use a led monitor with vga cable.

There are important things to mention:

- I have made an overclocking for gpu few months ago..but unfortunately I can't remember the exact values but I thinks they were 1120 for core clock & 1750 for memory clock & it was stable till it crashed 2 times after few weeks of overclocking & I thought that it was strange as if the overclocking wasn't appropriate why I haven't experienced artifacts at the beginning.. anyway I have lowered it to 1100 for core clock & 1650 for memory clock & the system was stable till the strange crash with grey screen happened one time while I was playing COD WWII & I had to make a hard reset for pc but it didn't happen again till I had these recent constant freezings am experiencing now.

- I have tried another gpu of a friend which is GTX 750Ti & games worked normally without any crashes so I believe there is no problem with ram, cpu or motherboard & the problem is with gpu or psu as gtx 750ti needs only one 4pin connector & on the other hand my r9 280x needs two 8 oin connectors so in my opinion it's not a solid experiment that I can determine that the issue is about gpu only

Finally, the problem is that I can't try my gpu in another rig that can handle it as I don't have any friend that has a solid psu..so if there are any other solutions that can fix my problem or at least to determine whether the problem is from gpu or psu or another component, I'll be so grateful & thanks in advance.

 
Solution
Slightly underclock both core (-50) and memory (-100), max power limit, increase voltage slightly.
Usually underclocking while supplying more voltage fixes these issues.

Finally, the problem is that I can't try my gpu in another rig that can handle it as I don't have any friend that has a solid psu..so if there are any other solutions that can fix my problem or at least to determine whether the problem is from gpu or psu or another component, I'll be so grateful & thanks in advance.
PC repair shops usually have test benches. I recommend you test it on another PC to rule out every other part you have. From cables to software and etc.

Also, it uses about 200W on average so a 450W+ PSU can power it...

omarmoustafa280x

Reputable
Apr 6, 2018
12
1
4,515
thanks for the quick response Kashimi, so do u think gpu is about to die or there is a problem that can be fixed ? ... another thing to mention I don't know if it's helpful but while am using unigine superposition benchmark , gpu performance is normal & am getting high fps as the utilization is 99% till pc freezes of course ):
 

omarmoustafa280x

Reputable
Apr 6, 2018
12
1
4,515
The strange thing is that I play on PC for only 5 to 7 hours a week so am not making heavy loads on it....maybe this is the reason why it did this as it needs care & more time to spend with :3 "am just kidding"
 

zebarjadi.raouf

Commendable
Jul 10, 2018
862
2
1,310
Slightly underclock both core (-50) and memory (-100), max power limit, increase voltage slightly.
Usually underclocking while supplying more voltage fixes these issues.

Finally, the problem is that I can't try my gpu in another rig that can handle it as I don't have any friend that has a solid psu..so if there are any other solutions that can fix my problem or at least to determine whether the problem is from gpu or psu or another component, I'll be so grateful & thanks in advance.
PC repair shops usually have test benches. I recommend you test it on another PC to rule out every other part you have. From cables to software and etc.

Also, it uses about 200W on average so a 450W+ PSU can power it.
https://www.techpowerup.com/vgabios/151600/sapphire-r9280x-3072-131205
 
Solution

zebarjadi.raouf

Commendable
Jul 10, 2018
862
2
1,310
10-25% voltage boost.

You sure none of your friends have 450W+ PSU? You could test with them.
 

zebarjadi.raouf

Commendable
Jul 10, 2018
862
2
1,310
I have raised voltage from 1200 to 1215 & now it keeps crashing in desktop...& lowered core clock & memory
So voltage boost + underclock didn't help. A test on another pc will tell if you need a GPU upgrade or not.

You can reset the clock and voltages if you haven't done so.

I'll search for these kind of pc repair shops
Any decent repair shop should have the parts ready. They can't repair stuff without testing them.
 

omarmoustafa280x

Reputable
Apr 6, 2018
12
1
4,515
zebarjadi.raouf it looks like underclocking has made it. I made a reset for voltage but I wondered what would happen if I decreased all the values except voltage to the lowest value, I made it for core clock(510 instead of 1020) , memory(750 instead of 1500) & power limit(-20 instead of 0) & temp is maintained at 60 Celsius... I have launched superposition benchmark twice & no crashing .. playing assassin's creed origins for half an hour & thank god no crashing .. so I think it's either psu can't feed gpu with the power it needs in full load or the gpu can't afford high frequencies any more due to dust, problem in thermal compound or sth. else I'll try to figure it out after I try the gpu on stock frequencies on another rig to make sure it's gpu or psu issue & until that I'll try to slightly raise core clock & memory until it crashes again ... so I really appreciate your help & thanks all for everything & I'll update with u when I try gpu on another rig.
 

zebarjadi.raouf

Commendable
Jul 10, 2018
862
2
1,310
I made it for core clock(510 instead of 1020) , memory(750 instead of 1500)
That's a big underclock.
so I think it's either psu can't feed gpu with the power it needs in full load
If I remember correctly, Just lowering power limit should lower power usage.
gpu can't afford high frequencies any more due to dust, problem in thermal compound or sth.
Thermal compound helps by lowering temps. If you got 85c+ while gaming or benching before, that could have been the reason. It's a good habit to keep everything in your system cooler than 80c to extend life.
I'll update with u when I try GPU on another rig.
If it artifacts on another system with healthy PSU and working drivers, I would guess 280X is starting to die. If not, mostly like it's your PSU.

Good Luck.
 
Status
Not open for further replies.