Memory/Driver Crashing

[TwilightPhoenix]

I've no idea if this is the correct sub-forum to post in but here we go. For months now I've been dealing with an issue in which my display driver will crash, causing my screen to go black for a few seconds and killing whatever application I'm running with a display driver recovered error. I've tried pretty nearly all the software fixes I could find. Lately is has been suggested to be that the issue might be improper or low voltage on my memory. To correct it, I've been told to set the DRAM voltage to 1.65 and see if the system stabilizes and the crashing subsides. I'm running an ASUS M5A97 motherboard, and DDR3 1600 RAM. Basically, my concern lies in the safety of this. Is it safe to do? Am I going to damage my CPU or my memory controller?

Thanks for reading.

 

[juane414]

I wouldn't increase your DRAM voltage until you are certain that your system is unstable because your RAM needs more voltage. Also, there is no specific number that you should set your DRAM to that will fix any problem. What you should do is download and run Memtest to make sure that your RAM is not defective. It your RAM passes Memtest without errors then it probably isn't a hardware problem.

While you run Memetest (just google it) can you please tell us more information about the specific hardware that you have? What graphics card do you have, and what RAM?

 

[TwilightPhoenix]

Of course, thanks for the reply. I'm running a GeForce 560 Ti. I'm shamed to admit that my system was prebuilt, and I don't know the exact make of RAM I'm running unless I pop her open and check. Is it that dangerous to overvolt, even by such a small margin? I've been told otherwise, though I need to be absolutely certain it'll not do damage.

Additionally, I do not have spare CD's on me at the moment so I cannot run MemTest just yet.

 

[juane414]

You can run Memtest off of a flashdrive as well. Part of the reason I was wondering what specific RAM you have was to check what voltage it's designed to run on. A lot of RAM nowadays is designed for 1.5v so 1.65 sounded high to me. If your RAM is unstable it shouldn't take much voltage to make it stable. However, the safer way to test if your RAM is clocked to high is to turn down the multiplier a step. Slower your RAM down will not hurt it. Overvolting it has the potential to hurt it, though the risk is very small if your overvolting only a little.

It is hard to help you out much without knowing more about your system. If your system is prebuilt from Dell or HP or some other company then you should be able to look it up online and find the specs. Did the GeForce 560 Ti come with the computer or did you add it in yourself? It could be a driver problem, it could be a hardware problem, it could be that your power supply can't give your system enough juice, or it could be any number of things. Try to give us more info about your system if you can.

 

[TwilightPhoenix]

Not much, you say? Well. It's running at 1.500v presently. I did bump it up very marginally, 1.50645v. It crashed twice within three hours. I'm not sure if that's a sufficient amount of juice to make it stable or not, but I did try. But I reset it back to auto-clocking the moment it did happen to crash.

The system was built by iBuyPower. I have my system invoice from the day I placed the order, but it's rather vague and doesn't offer specifics. The only way I'm going to get the specific information on the RAM is by opening it up and pulling a stick. As for the specifics of my system:

ASUS M5A97 mobo
GeForce 560 Ti
8g's of DDR3 1600 RAM (you knew that, of course)
700W PSU (generic, in-store PSU. I know, I know.)
AMD 1100T hexcore CPU

The video card came with the build, though it has been a pain. For instance, it'll not cool properly sometimes. The temp will hang at 50-51c even though the fan is going, and I'll have to reboot to get to cool back down to the idle 36c. I've been trying multiple drivers, slowly working my way back to the older stuff as each fails me. I've tried a complete reformat in the past to no avail, granted I did install drivers. In terms of it being a weak PSU, the driver crashing and the aforementioned heat thing have been my only noticeable issues with it. No odd shut-downs, no other failures of any kind that I've noticed to be problematic.

 

[juane414]

Well this sounds like a problem with your GPU, not your RAM, but running 1.5v RAM at 1.65v is probably not a good idea. That's a significant voltage increase and you could damage the RAM. So definitely don't do that.

Since you reformatted and reinstalled drivers, and that did not fix the problem, it seems unlikely that this is a problem with the graphics drivers. I did some googling and it seems like other people have experienced this problem. Try downloading the latest DirectX software from Microsoft and update Adobe Flash Player to the latest version as well. Also, go to the manufacturers website for your graphics card, look up your model, and see if there are any updates for it. There might be a new bios update for your graphics card. If you're not sure what GPU you have just download GPU-Z and it will tell you.

 

[TwilightPhoenix]

Isn't updating a GPU BIOS a massive procedure? I toyed with that idea very briefly but I've seen a lot of warnings about it.

 

[juane414]

It can be a massive procedure, but it can also be a simply procedure. It really depends on what GPU you have and which manufacturer it is. Each manufacturer has a different way of doing it, but if you follow their directions carefully you shouldn't have a problem.

 

[TwilightPhoenix]

There's one thing I forgot to mention. It seemed like putting my system into power-saving mode in BIOS helped alleviate some of the crashing. I suppose that could be a symptom of the PSU, but I'm not entirely sure. There's another thing I was curious about. Throughout all these months, there's been one constant in terms of software. MSI Afterburner. Could this possibly be the root of driver crashing? After all, it does interact with a GPU on a nearly constant basis.

 

[juane414]

What are you using MSI Afterburner for? If you are using it to overclock your GPU then your overclock could be causing the crashes. Unstable overclocks cause crashes all the time.

 

[TwilightPhoenix]

There's no overclocking on my system. I'm using it for the custom fan profile. When I first bought my system, my GPU would run in the 60's under heavy load. In Skyrim, for example, I'd rarely go above 68c. Now I'm consistently in the mid 70's to low 80's with an aggressive curve. Without it I'm not sure how high my temps would reach.

 

[juane414]

With those temperatures its possible that your GPU is overheating. Over time fans and computers build up with dust and restrict airflow. It could be that the inside of your case needs a good cleaning. Have you made any modifications to the computer since you got it that might account for the rising GPU temps?

 

[TwilightPhoenix]

I've cleaned the fan and metal of the heatsink of my GPU as best as canned air will allow. Naturally there's dust that hangs onto that metal with a vengeance and it never completely cleans off. I've made no hardware modifications beyond a PCI-e fan thing. It sits two slots underneath the video card, and I hoped it'd provide cooling. It doesn't. Next time I deep clean I'm going to move it directly under to see if that lowers temps but haven't yet in the off chance it actually increases them.

 

[juane414]

Maybe try running your computer with the side panels open and see if your temps change. That's a good way to find out if you have adequate airflow in your case.

The best way to find out if your GPU is causing problems would be to swap it out with a different GPU. If you have a spare GPU lying around you could try that. Another thing to try would be to download 3D Mark Vantage or 3D Mark 11 and run that a few times. If your computer crashes during the test then your GPU is probably overheating or having some other problem. If your computer is crashing randomly when the GPU is not being used much (like web browsing) then it's more likely to be a RAM problem.

Also, since you bought your PC from iBuyPower... have you tried contacting them about this problem?

 

[TwilightPhoenix]

The case is a Coolmaster HAF. One whole side of the case is basically open holes with two fans, one on the top and another in the front. It's probably not a cooling issue. Unfortunately, there's no spare GPU's around otherwise I'd have tried that in heartback months back. As for the crashing, it's completely random. Could be gaming, or could be web surfing. Hell, it crashed the other day typing a response on these forums.

I haven't contacted iBuyPower, it'd probably be a pointless endeavor. Unfortunately my parts warranty is up, so they'd offer me no help that I wouldn't have to pay for first.

 

[juane414]

Well, honestly, if the crashing happens randomly it's more likely to be a problem with your RAM than anything else. My cousin just had a similar issue with his system. It would crash during games, during web browsing, even during booting or restarting. We ran Memtest and got tons of errors. So we removed one stick of RAM and ran Memtest, and then swapped for the other stick of RAM and tried again. One of the RAM sticks passed Memtest just fine and the other one through up all kinds of errors. So it turns out that he had one bad stick of RAM.

Unfortunately, when you have problems like this the only thing you can do is trial and error. You just have to keep testing things and eliminating possibilities.

Here are some things you can try:
1. Run Memtest
2. Run Memtest on each stick of RAM individually
3. Monitor your temps carefully with HW Monitor and make sure nothing is overheating.
4. Try running your system with minimum components (remove/unplug any unnecessary peripherals and just run some tests PSU, CPU, heatsink, Mobo, RAM, GPU, HDD)

 

[TwilightPhoenix]

How do I run MemTest off my flash drive, and how long do I allow it run?

 

[juane414]

Get a USB drive that doesn't have any data on it, or save the data to your hard drive first.

Format the USB drive (start>computer>right click USB drive>format)

Download the Memtest USB drive auto-installer.

After it downloads, double click the file and run it. Just follow the instructions and Memtest should be installed on your USB drive.

Restart the computer and go into the BIOS (usually you have to press the delete key during boot up).

In your BIOS navigate to the boot order section and set your first boot device to your USB drive (usually it's called USB_HDD or something similar).

Then save and exit the BIOS (leave USB drive in the computer).

When your computer restarts it should boot right into Memtest and start running tests automatically. If you have defective RAM then you should start getting errors within a few minutes. I would let Memetest run for 20-30 minutes to make sure there are no errors.

 

[TwilightPhoenix]

Running MemTest86+ now. I'll get back to you in a few hours.

 

[TwilightPhoenix]

Right. I allowed it to eight hours and make five passes. I returned zero errors on RAM. I think it's a pretty safe assessment to say the problem isn't my memory.