New System Problems

[Torric]

I just built a mid-range gaming rig from newegg, and it is awesome. But when ever I try to run games, they play for about a minute before freezing w/obnoxious noises and I have to reboot to get out of it. Internet and other tasks work fine though.

Intel Q6600
MSI P43 mobo
EvGA 9800 GTX+
4GB GSKILL 1066 DDR2 RAM
WD Black 500GB HDD
Raidmax Smilodon Mid-Tower Gaming Case
OCZ ModXStream Pro OCZ500MXSP 500W ATX12V V2.2 / EPS12V SLI Certified CrossFire Certified Modular Active PFC Power Supply

 

[Proximon]

You could start by getting your CPU temps:

http://www.techpowerup.com/downloads/1139/Real_Temp_2.70.html

 

[Emperus]

Disable features like Enhanced Speed Step from bios.. In case the problem persists then the temp issue might be present.. Try keeping your case door open and then launch the game..

 

[dextermat]

What game do you play?

Usually when you get there problems:

1) processor overheat (69-70 degree)
2) video card overheats (depends on models 85-90 degree)
3) Chipset on motherboard might overheat (are their fans / heat sinks)

Donwload speedfan and look at thoses temperatures

Have you build your own system, might be the termal paste on the processor or the fan might be installed correctly

good luck

 

[Torric]

I have the fan on my gpu running at 50-75% usually, so unless it is internal or the PCI-E slot is messed up, I don't think it is the graphics card. My case has three additional fans, but I am only able to run one atm because I have one less cord than I need to divert power for all three. I'll download that cpu app and let you know what I found. Thanks!

The games (the only ones I've installed so far) are CoD World at War and Neverwinter Nights 2. They were on almost max settings, and they ran fine until they crashed...

 

[Torric]

So it seems that if I start CoD 5, and alt-tab out of the game, it crashes by itself and does not affect me like it would if I were playing the game. I discovered this while I was watching the cpu temps. The temps only reached 44 C, before it crashed.

Edit: I tried the test again, this time alt-tabbing in and out of the game, and I was able to play a lot further than I had when I just straight played. The cpu temps dropped when I went to the desktop, and the most recent stats were 56+ before I alt-tabbed back into the game before it crashed. I suspect that the temps rose past the threshold, but I do not know what this number would be. Would anyone be able to tell me please?

 

[Zorg]

That's not how you check your temps. Run Prime95 small FFTs and monitor your temps with Core Temp.

Run prime95 for at least 15 minutes, or until temps stabilize, and post the results. If the temps hit 80C then stop Prime95.

 

[Proximon]

If you had downloaded real temp, it would have kept track of the maximum temp for you.

 

[Torric]

RealTemp is what I used to monitor my cpu. I just could not get teh final temperature because I could not get back to the desktop when the game crashed, and the temps were not saved for me to find when I re-booted. I guess I'll run prime95 and see what happens.

 

[merlinbadman]

What graphics drivers are you using?

COD5 does have crashing issues in general, my mate is always complaining of it.

 

[Torric]

Well, it also crashes with NWN 2. And probably any other game as well. I am using the drivers that the EvGA disk provided for me. I'm also going to run Prime95 today so I'll post the results.

 

[Proximon]

It's not hard to do. Quick download and set up, and you'll have a pretty clear idea in about 60 seconds if it's a problem.

 

[Torric]

So I'm in the BIOS, and I went to PC Health Status. Where it has the voltages 3.3V 5V and 12V, should the corresponding voltages be lower, higher, or in the general area of those? Just wondering.

 

[Proximon]

They tend to not be very accurate. You really can't depend on those readings at all, but low voltages listed there might be enough to make you look harder at the PSU, maybe, if you were having problems.

 

[Torric]

So I did prime95, and after a minute, it shut off and said FATAL ERROR: Rounding was 0.5 expected less than 0.4. And it said this for all of the cores. Next it mentioned detected hardware failure and then it stopped the test. What do y'all make of that?

 

[Proximon]

Did you test using memtest from a boot disc? I loose track on these threads who did what

http://www.memtest.org/

 

[Torric]

No, I downloaded Prime95 and did the middle torture test. Tomorrow I am also going to remove and apply some thermal paste to the cpu cooler to see if it changes anything.

 

[Proximon]

If Prime95 errored out before you saw any jump in the temp, I doubt that's an issue at this point.

More likely motherboard or RAM.. beyond that PSU comes in 3rd.

 

[Torric]

How would I check to see if it is the mobo or the RAM? And are certain mobos compatible with only certain types of RAM?

 

[Proximon]

The link I gave a few posts back will get you what you need to test memory.

Well, certainly various chipsets support different kinds of RAM... but your P43 board supports DDR2. Whether it configured correctly to run it properly, I can't say.

A memory test should come first.

 

[Zorg]

Real Temp and Core Temp both have temp logging. As Proximan said, it is probably a RAM issue not a temp issue. Unfortunately you never posted the temps when it crashed so who knows for sure.

What were the temps when Prime95 crashed?

Run memtest86+.

If you see red, then run each stick separately to look for a bad stick.

 

[Torric]

Ok, I'll get to running the memtest today. I'd be happy if it was only a RAM problem, because it is a lot cheaper than a new cpu...The cpu temps were between 30-40C so nothing too high.

 

[Torric]

So I've tried both sticks, seperately and together, and no test got past 51% complete. The test didn't pick up any errors, the PC just froze. It also shut down by itself when I tested both sticks with 4-test of memtest86 running like it told me to. Would this be a problem with the DIMM slot? Or does this signify some other problem, such as both sticks are faulty.

 

[Torric]

Well, I just did another memtest86 test for each stick and it said ERROR- your computer cannot accurately store data RAM, you need to fix this, etc. So unless their is a way to check to see if it is the DIMM slot causing these issues, I'm going to get a refund from Newegg and buy a different brand of RAM.