Sign in with
Sign up | Sign in
Your question

Brand new PC crashing (Not OC'd)

Last response: in Motherboards
Share
August 14, 2010 12:54:05 AM

Mainboard: MSI 890GXM-G65 (http://www.newegg.com/Product/Product.aspx?Item=N82E168...)
Processor: AMD Phenom II X4 955 (http://www.newegg.com/Product/Product.aspx?Item=N82E168...)
Memory: G.Skill DDR3-1033 (http://www.newegg.com/Product/Product.aspx?Item=N82E168...)
Gpu: GeForce GTX 460 (http://www.newegg.com/Product/Product.aspx?Item=N82E168...)
PSU: Tuniq 650W (http://www.newegg.com/Product/Product.aspx?Item=N82E168...)

This computer is 4 days old. I have done a lot of miscellaneous testing to try and troubleshoot what is causing errors I am having. It is running all stock settings (BIOS is latest version as well).

Background: I installed SC2 a day after the PC was up and running (seemingly okay) after playing SC2 for maybe 30 minutes I got this crash
Quote:
ACCESS_VIOLATION (0xC0000005)
occurred at 0023:00C9C9F7. The memory at '0x00000018' could not be read.


I looked into this crash quite a bit, most people pointed at memory being faulty, some said to tinker with the memory voltages (lost me at that).

Now that I mentioned voltages, I should also mention that I have never touched voltages, and never really have had any interest in touching them. I have never overclocked anything before therefore if I am slow when it comes to some terminology, don't scream :( 

I have tested each individual memory module 1 at a time. I haven't done it for more than maybe 2 passes each but as you can imagine I am kind-of anxious to get this working for I have gotten literally 0 play time out of it thus far. So as far as memory being BAD, I don't believe this is the cause.

After some extreme googling I ventured into a website that talked about benchmark testing programs/etc, I downloaded Prime95 and ran the stress test (blend) on my computer and this is what I got soon after:

FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
FATAL ERROR: Rounding was 0.49609375, expected less than 0.4
Hardware failure detected, consult stress.txt file.
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
FATAL ERROR: Rounding was 0.486328125, expected less than 0.4
Hardware failure detected, consult stress.txt file.

so I ran it again just to be sure and got this:


FATAL ERROR: Final result was F7E55DB9, expected: 61182009.
Hardware failure detected, consult stress.txt file.
Self-test 1024K passed!
Self-test 1024K passed!
Self-test 1024K passed!

I then tried the "smal FFT's test" and got this:

Self-test 8K passed!
Self-test 8K passed!
Self-test 8K passed!
Self-test 8K passed!

What does any of this mean? I have no clue, but hopefully someone here will!

Other oddities thus far:

Sometimes, Windows Aero Effects will just randomly disable. My screen will flash, and then the glass-like transparency will be gone. (My video drivers are up to date, I have uninstalled and reinstalled them 3 times)


Sometimes when I am in Firefox and open a tab.. the entire screen will freeze, then I will attempt to run task manager and everything will delay for a minute or so, then everything will pop up at the same time. I have seen this in XP before, but this is a brand new install of Windows 7. Similar events happen in other random applications, such as Steam (when launching initially).


My temperatures according to Speedfan:

GPU: 36C
Core: 35C

Both the GPU and Core maxed out at 60C when running the stress test, but never went higher than 61.


I am seriously ready to just return the CPU/Motherboard/Memory and replace it with three different/new ones.

More about : brand crashing

a c 435 V Motherboard
a c 128 } Memory
August 14, 2010 1:39:14 AM

You may just have to pay the restocking fee with most returns and get another brand of ram. I suggest you try finding compatable ram on Msi's website for your board. I like crucial or corsair the best, but also run pny and micron with no problems. the other alternative is trying to check your current bios file listed on the post screen, and compare it to the bios file on msi's website. If the latest bios file doesn't match yours, you can flash the bios using msi's "live update". To do this, you have to sometimes disable "boot block" in the bios, and any firewalls in windows or with virus protection software. Then download the live update software off msi's website and install it on your pc. After installation, you go online back to msi's website and run the live update program. If it runs successfully, you will get a small black window that starts off erasing your current bios, then loading the new one and checking it before rebooting the system. You must not shut down the computer while flashing, or the board will be dead.
m
0
l
August 14, 2010 2:07:16 AM

I appreciate the response.

On MSI's website they have listed "•Supports four unbuffered DIMM of 1.5 Volt DDR3 800/1066/1333/1600*/1800*/2133* (OC) SDRAM, 16GB Max "

The memory I have is unbuffered 1.35V 1333, would this really make a difference?
m
0
l
Related resources
a c 435 V Motherboard
a c 128 } Memory
August 14, 2010 10:23:22 AM

No, it shouldn't. Check the default ram voltage setting in the bios. Even if it's .1 volts too high, it should still run. You can also rma the ram to gskill and avoid the restocking fee, but you may still have the same problem with the new ram. That's why I recommend a different brand.
m
0
l
a c 156 V Motherboard
a b } Memory
August 14, 2010 1:23:22 PM

Joines said:


After some extreme googling I ventured into a website that talked about benchmark testing programs/etc, I downloaded Prime95 and ran the stress test (blend) on my computer and this is what I got soon after:

FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
FATAL ERROR: Rounding was 0.49609375, expected less than 0.4
Hardware failure detected, consult stress.txt file.
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
FATAL ERROR: Rounding was 0.486328125, expected less than 0.4
Hardware failure detected, consult stress.txt file.

so I ran it again just to be sure and got this:


FATAL ERROR: Final result was F7E55DB9, expected: 61182009.
Hardware failure detected, consult stress.txt file.
Self-test 1024K passed!
Self-test 1024K passed!
Self-test 1024K passed!

I then tried the "smal FFT's test" and got this:

Self-test 8K passed!
Self-test 8K passed!
Self-test 8K passed!
Self-test 8K passed!

What does any of this mean? I have no clue, but hopefully someone here will!


We use Prime95 to stress test an overclocked system.

It will load all the cores of a CPU with a 100% load. That does two things. First, it will heat the CPU to its maximum temperatures in about 10 minutes. And second, it does a detailed mathematical check of the processing in each core.

If you select the "small fft's" test, you primarily test the CPU because the small fft's test runs in the CPU cache. If you select the "large fft's" or "blend" test, you are moving enough data to and from the memory to really wring the memory out.

Failing the "large fft's" or "blend" test indicates a problem with memory - but you knew that. :) 
m
0
l
August 15, 2010 12:33:50 AM

Thanks for the replies everyone. I ended up testing each individual memory module for at least 6 passes (and one for 26 passes over night). Turns out that two of them began to have errors around the 4th and 5th passes around 85 or so errors for one, and about 5 for the other.

Went out to a local computer store we have here and bought some corsair memory, put it in and ran Prime95. For the first time since I've had this computer, Prime95 passed the blend test the first time around. It normally would crap out after Test#2 or even Test#1

G.skill = blah

2 out of 4 brand new modules defective?

Next time I'll spend the extra $20 for memory thats actually good.

Thanks again!
m
0
l
!