AMD Card Drivers Causing Crash

[OrigD]

As the title suggests I believe that the AMD drivers for my graphics card are causing my computer to crash a few minutes after startup. Regardless of what I am doing, gaming, browsing or even just leaving it idle it will either crash/freeze/bsod. After it has crashed a few times it starts up with only 4GB's of ram usable (according to bios and computer properties) as opposed to the 8GB's installed at which point it usually runs *fine* in the sense that it wont crash, however it will be far slower than usual. I have already gone over this and it would seem there is no issue with regards to the ram.

I have attempted having a clean install of windows however, after a short while the problem always seems to return, if I reinstall my AMD drivers it will work for that startup, once I power it down and power it back up the problem will come back as well. I have tried to downgrade to specific versions of drivers such as 15.11.1 which is supposed to fix certain crashes however this has not helped.

I will provide a memory dump in the hope that someone may be able to make better understanding of it than me as I haven't the slightest idea. *Turns out im not too sure how to provide the memory.dmp*

Many Thanks

AMD FX 4100 - Quad Core Processor
R9 380 4GB
8 GB's Ram (2x4GB Sticks)
500W Corsair PSU
Asus M5A 78L - M LX

 

[10tacle]

After it has crashed a few times it starts up with only 4GB's of ram usable (according to bios and computer properties) as opposed to the 8GB's installed. I have already gone over this and it would seem there is no issue with regards to the ram.

Does that mean you have tested each memory module individually and/or you have run a memory test program on each module (and both at the same time) and there were no errors? The fact that it was only reading 4GB is pointing to something unrelated to your GPU drivers. I would first look at all hardware before looking towards software. That includes the power supply and making sure it's not failing (dipping below +/- 5% 12v and +5v ATX specs specifically).

Also, are your 2x4GB sticks from a 2x kit or were they purchased as individual sticks?

 

[OrigD]

Yeah tested both sticks individually, as well as the slots that they were in. Also did them both at the same time as well, no errors came up. Both sticks were purchased together.

How do I test the PSU, I have had suspicions about it since sometimes the computers anti surge protection gets triggered, however, as I said it seems that no issues occur if I remove the AMD drivers.

 

[10tacle]

Both sticks were purchased together.

I assume that means a kit as in they were in the same package and not individually packaged (meaning non-kit form).

How do I test the PSU, I have had suspicions about it since sometimes the computers anti surge protection gets triggered, however, as I said it seems that no issues occur if I remove the AMD drivers.

Several ways. The two I use are HWiNFO64 and MSI Afterburner when running a CPU/GPU stress program (AIDA64, OCCT). If any of your voltages are outside of the +/- 5% ATX spec, then your PSU is likely failing. Here are the parameters:

+12v -> 11.4v to 12.6v
+5v -> 4.75v to 5.25v
+3.3v -> 3.135v to 3.465v

 

[OrigD]

Currently out at this moment, as soon as I get back will take a look at the PSU.

 

[OrigD]

Well I started up the HWiNFO and after a short while the computer crashed as expected, luckily it froze on the screen with the voltages however they appear to be completely fine. No values dipped below or above at either their maximum or minimum.

 

[10tacle]

Okay that rules out the power supply and the memory as hardware. Have you tried rolling back to older Catalyst drivers? I don't think that's your issue. I still think it's hardware related some where. Either a failing CPU (less likely) or motherboard (more likely).

I checked reviews on both Amazon and NewEgg of that board and 10% of the Amazon owners and a whopping 29% of the NewEgg buyers reported either severe problems or failure at some point. Since that chipset is six years old now, I would strongly suspect you have a dying motherboard. I had a dying ASUS P8P67 Pro for my i5 2500K build. It only lasted two years before blue screen crashes and just complete shutdowns with no warning. I replaced it with an MSI Z77 board and was back up and running.

You can try flashing to the latest BIOS but that may cause more harm than good and possibly brick the board entirely if it's on its last legs.

 

[OrigD]

The company that produced the GPU offered updating my motherboard to its latest version as a possible solution as well as trying to downgrade to other versions such as 15.11.1 (many other users also recommended) however none of those seemed to provide me with the fix I needed. As for the motherboard I don't suppose that there is any way to really test as to whether or not it is failing?

 

[10tacle]

As for the motherboard I don't suppose that there is any way to really test as to whether or not it is failing?

Unfortunately no. Not without professional equipment like a digital multimeter and knowing exactly what you are looking for and doing (that's what refurbishing companies do contracted to motherboard makers who fix and re-sell RMA'd boards to Etailers like NewEgg as refurbs). You might be able to spot something by the eyeball like a blown/burned out capacitor but if you don't know how to repair electronics, it wouldn't matter anyway. It could be a failing chip component like the memory controller or something in the CPU bridge that you can't see.

If I were in your shoes, at this point I'd seriously consider investing in a new Ryzen build if you favor AMD. I would not put any more money into your current build. Your GPU is still a strong GPU and if your PSU is one of the newer Corsair CX lines, that's fine too (older CX series built prior to the second half of last year were crap).