System randomly freezes during usage, then sometimes on POST on reboot

[brazenwolfy]

Hi all,

In the past few months my computer has become prone to random freezes during Windows login screen, during normal usage, or while the computer is in sleep mode (it would power up randomly in the middle of the night and freeze). At this point, if I reboot my system, it will randomly freeze on the POST screen. It would take a few reboots to get back to loading Windows.

I really don't know how to troubleshoot this. If I look in the reliability history in Windows 7, the only issue it recognizes is that I didn't shut down Windows properly. Is there any way to better troubleshoot this? So far I was able to run memtest overnight with no errors and have run chkdsk on all harddrives attached to my system.

Here's my specs:
Motherboard: Gigabyte X58A-UD3R
CPU: Intel(R) Core(TM) i7 CPU 950 @ 3.07GHz (Bloomfield)
Memory: 12GB (6x2GB) OCZ3G1600LV2G
HDD: WDC 1TB
GPU: evga Nvidia GTX 470
OS: Windows 7 (64-bit)

Can post any other information or the CPU-ID report somewhere if needed.

Thanks,
BW

 

[Chainzsaw]

Hello,

What is your power supply make and model?

 

[brazenwolfy]

I'm using a Cooler Master Silent Pro Gold 800 Watts Modular Power Supply

 

[Chainzsaw]

Ok, I was going to say it might be a PSU issue however it looks like it provides plenty of power.

It sounds like a RAM issue - however you already did some memtest tests and you said it passed.

I'm inclined to think your RAM is not set up properly in the BIOS. Have you tried clearing your CMOS?

Also you might want to try re-seating your RAM as well.

Edit: You have 6 - 2GB sticks in your system as well? If so, all your DIMM slots are full right?

Edit 2: These are your exact sticks:

http://www.ocztechnology.com/ocz-ddr3-pc3-12800-gold-low-voltage-dual-channel.html

I noticed they are supposed to run at 1.65V, however your motherboard uses/requires 1.5V sticks in the DIMM sockets. I think that might be your issue - running 1.65V DDR3 sticks in a 1.5V motherboard. <--- what is your RAM Voltage set to in the BIOS?

Do you have another set of ram you could try out that has 1.5V ram?

 

[brazenwolfy]

Yes, all DIMM slots are filled on my system.

According to BIOS the voltage is 1.52V. Is there any way to increase this? Currently do not have any backup RAM to use.

 

[Chainzsaw]

There should be an option in your BIOS to register the ram at 1.65V. Do not set it any higher.

Edit: Heres your manual, click on downloads and manual

http://www.gigabyte.com/products/product-page.aspx?pid=3449#manual

 

[brazenwolfy]

I set the DRAM voltage to 1.64V in BIOS and the freeze still occurs. Even though I set the voltage manually, CPU-ID still reads the voltage for all RAM slots at 1.5.

 

[Chainzsaw]

Is your BIOS up to date? I would reccomend updating it via your BIOS/CMD prompt and try it again.

If that still doesn't work...im out of ideas besides buying proper 1.5V ram kits.

 

[brazenwolfy]

Currently using Award BIOS FB, 8/24/2010. There are newer BIOS versions listed on the website, but I'm unsure if I should update to them:

http://www.gigabyte.us/products/product-page.aspx?pid=3449&dl=1#bios

I attempted using the @BIOS software included with my mobo to update from the server, but that didn't seem to work. Would it be safe to update to version FH?

 

[Chainzsaw]

Currently using Award BIOS FB, 8/24/2010. There are newer BIOS versions listed on the website, but I'm unsure if I should update to them:

http://www.gigabyte.us/products/product-page.aspx?pid=3449&dl=1#bios

I attempted using the @BIOS software included with my mobo to update from the server, but that didn't seem to work. Would it be safe to update to version FH?

Your BIOS appears to be very out of date. I would reccomend updating to the FH if you feel comfortable enough to do it. However a few problems:

If you do update your BIOS, and it freezes during the update, it might bork your system.

The update might actually make things worse.

On the flip side:

It might fix your issues of your RAM voltage not sticking.

It might even fix everything.

Edit: One more thing, you have to decide whether or not you want to update your BIOS. Like I mentioned above, it could either fix everything, or you could bork your system. So update/don't update the BIOS at your own risk!

 

[brazenwolfy]

Hm, trying to update the BIOS right now and it said the BIN size isn't correct. Getting this error using @BIOS utility and the QFLASH utility.

 

[Chainzsaw]

What medium are you trying to update from? A USB key?

If it's a USB key - make sure its formatted in FAT.

Also when you downloaded the BIOS - was it the full size and not a garbage download?

 

[brazenwolfy]

I was able to flash my BIOS using the Gigabyte util @BIOS (had to reinstall this utility for it to take). In the BIOS I've set DRAM voltage to 1.64 V and manually set the timings to 8-8-8-24.

This doesn't appear to be reflected when I run the CPU-Z application under the SPD tab which lists voltages as 1.5V.

 

[Proximon]

The SPD tab is a set of potential voltages. Those are settings you COULD use, according to the data on the RAM modules.

What you should do for testing purposes is remove 3 sticks. Make sure the RAM settings match a set of settings from the SPD tab, and test that way.
If that does not resolve it then test with the OTHER 3 sticks, only.

The PSU is not great, it's an Enhance-built model that has a few minor issues. The amount of available power has very little to do with the quality of a PSU. I would not rule it out just yet.

 

[brazenwolfy]

Alright, thanks for the extra info. I'll try testing each RAM set individually when I have time this weekend. For now, the system froze again, and I was welcomed to the following message on the POST screen:

It looks like the RAM speed is incorrect. How would I fix this?

 

[Proximon]

Resetting to BIOS defaults should resolve that.

 

[brazenwolfy]

So after updgrading BIOS to the latest version and setting up my memory to the correct voltages and speeds, my system ran pretty smoothly for a few days. However, after a few flickers of a kernel crash on my NVIDIA card, my system froze once more. After a few reboots, my system froze at several points on the POST screen as well as the screen which prompts for safe mode. I had to completely power down my system and turn it back on for Windows to boot properly. Would this indicate an issue with the Mobo or HDD? Or is it more likely still a memory issue?

 

[brazenwolfy]

Some extra info:

Noticed my kernel driver for my NVIDIA GTX 470 card sometimes crashes and recovers a few times before my computer freezes. Is there any way the graphics card is causing all these issues?

 

[Proximon]

I still suspect the PSU. That's the one part that best fits the various symptoms at this point... no way to know without swapping it out.

It COULD be the GPU or the MB or other parts. Swapping out the GPU might help narrow things down, but only if it was exchanged for a card of equal power demands.