PowerColor R9 270 Red screen of death and BSOD

Crowe55

Honorable
Jul 3, 2013
5
0
10,510
Hello! First of all, here's my hardware:

CPU: Intel Core i3-6100
RAM: Kingston KVR21N15S8/4 (A single 2133 4GB DDR4 stick)
Mobo: ASRock B150M Pro4
HDD: WD Caviar Blue 1TB 7200rpm 64M
PSU: Fractal Design Integra M 550W (80Plus Bronze, 45.5A 12V amperage, tier 3, according to Tom's PSU tier list)
GPU: PowerColor R9 270 TurboDuo
Monitor: Samsung SyncMaster 720N (1280x1024)

Now, the problem. The explanation is going to be rather long, so you can check the TL;DR version after it.

I assembled the PC, fired it up, installed Win10 Pro x64, some software and a few games - GTA V, Rome II Total War and Crysis 3. I then proceeded to play some GTA, and everything worked fine until about 25 minutes into the game, when I got this BSOD: Click

Restarted, opened GTA V, and again after about 20mins i got a BSOD, this time with a different error code: Click

Reinstalled GTA, fired it, played a bit, then this BSOD: Click

Restarted, launched Crysis, played a bit, and this time I got a Red screen that stayed for a minute and then the pc rebooted: Click
[Note: The blue line on the monitor is a known issue - dead pixels]

I then downloaded Memtest x86, and ran it: Click
It froze at 98% first pass for 2 hours.

I restarted and ran Memtest again with this result: Click and later again, with this result: Click

I then changed the RAM module's position from slot A2 to A1 on the motherboard. Blue screens stopped, red ones - didn't.

Then I reinstalled Windows 10 by wiping the C drive, or in other words - a clean install. Red screens persisted.

Checked the GPU's 6-pin power cable, replaced it with another one. Downloaded FurMarks to stress test the GPU - ran it for 2 hours - no screens, no bsods, no rsods. Max temperature reached 68 C, max fan speed - 1200rpm as reported by GPU-z. I never overclocked the
card. The red screen while gaming persisted.

Took out the card, and ran on the integrated Intel HD 530 graphics. No BSODs or RSODs while gaming, although GTA V seemed to "Dont send" error me a few times.
Put the card back in - Red screens came back as well.

Downloaded Guru3D's Graphics driver uninstaller, uninstalled AMD's 15.7.1 driver, and installed it clean - red screens persisted. Tried with the lates beta driver - 15.10, red screens persisted.

Switched the OS to Windows 8.1 pro x64, and did all the same procedures and got the same results. I believe this means I'm not dealing with a poorly optimized driver for win10, and that my problem isnt OS related at all.

Currently I'm running 8.1 Pro x64 and I managed to get a few minidumps from the red and blue screens, and the links to them are here: click

I intend to try one last thing before i give up - wiping the whole hard drive and proceeding with a UEFI Windows installation.

TL;DR - I believe my graphics card or RAM may be defective. I don't think my PSU is too weak for this build. My GPU's and CPU's temperatures are within the safe range. So what should I do?

To anyone who read the whole post: Thank you!

EDIT: Fixed the links to the images



 

Crowe55

Honorable
Jul 3, 2013
5
0
10,510


I thought about doing so too, however unfortunately, the only other PC i have that I can use for comparative testing is a Jurassic-era Core 2 Duo E4500-based system with a very old motherboard that has PCIe gen 1 slot, pretty aged capacitors and a crappy no-name PSU. Of course i suppose i could move both the GPU and the new PSU to that old system, but I'm afraid that that might do something bad to the new parts. Please correct me if I'm wrong about this.
 

Crowe55

Honorable
Jul 3, 2013
5
0
10,510
Nevermind, I discovered the culprit.

Turns out the stock BIOS version that my motherboard came with, had some RAM compatibility issues.

I updated the BIOS to the latest P1.40 version and voilà - no red, blue, green or whatever other color of the rainbow screens whatsoever.

Now, how can I mark this thread as "solved".....