Hard Crashes ( help )

[Lyonheart]

Not sure where to post this.. so forgive me if this is the wrong place for it. I have been getting hard crashes last few days. semi reliable crashes.. i mean that they happen for sure If I'm multi boxing. Ill post my PC specs at the bottom and anything that might help. Ill try to describe what happens best i can. When i crash the screens ( using two monitors ) get crazy lines all across them and then the PC just shuts down or reboots. out of the 10 or so times it has done this i have got a glimpse of a critical error box once.. but most the time it just crashes and never has time to let me read any error message. So I'm a bit clueless as to why its happening. I get no blue screens. just BAM!.. reboot.

It seems to only happen when I'm multi boxing. I have played for hours on my main with no crashes. It started the last few days ( SP1 maybe? ) I had a crash problem one time on another PC that ended up being my sound blaster card drivers or something.. disabling sound fixed it ( boo ) I disabled sound this time and it still crashes. so its not that.

my specs are

P Q6850 3.00 OCd to 3.66 ( water cooled )

1k wat coolermaster PSU

Asus P5E3 deluxe/wifi lga 775 x38 MB

4g of ddr 3 mem

Vista ult64 bit ( SP1 )

two rad 3870s in xrossfire

1TB hitachi 7200 HD

If its SP1 ? is it hard to revert back to the last version ?

Any help or advice . i will be gratefull ( before i have to haul this big ass PC to the shop and be without my game for days 8( ..hehe )

[keyclone]

could be a couple of things... off the top of my head:

1. in-game, make sure you have 'sound in background' UNCHECKED (could be a sound driver issue)

2. make sure you have the latest sound drivers... or eliminate the problem by removing the sound board or disabling sound all together

3. could be faulty memory

4. could be a faulty power supply

5. 'dirty' power... are you connected to a battery backup? it would condition the power and handle any 'hiccups' from the power company which could result in your machine 'browning down' enough to trigger memory issues and a reboot

6. could be a cracked motherboard (only ever seen this once)

7. a problem with some kernel level driver causing the reboot (i see this as unlikely since the screen does funny things before reboot.. driver issues would go straight to reboot, unless the faulty driver is the video driver)

8. a general short on the board (good luck with that one... unlikely tho)

that's what i've got... maybe something will trigger a thought and help you solve it.

good luck

[Lyonheart]

thank you.. I always hate these mystery crashes.. seems like it takes forever to pinpoint them 8(

[Lefrog]

Never had any experience with your motherboard, BUT

I've personally had a lot of problems with any kind of ASUS motherboard while trying to dual vid card/SLI it up including crashes like you described. I would try taking one of the video cards out and switching the motherboard to single mode (there should be a little chip next to slots that you can switch around). Try to find a way to set it up so you can box without both of em and try it out and see if you have any crashes.

Hope this helps =)

Edit: Im almost sure its the motherboard, i just checked one of my rigs that seems to have same problem and i have a ASUS P5NSLI in it. I've been running tests of on the different hardware parts and they all seem ok. Its always the hardest to diagnose if its the motherboard =(

[Lyonheart]

thx again for the tips. ill try removing one video card. It seems this started happening soon after i installed SP1 and the new crossfireX drivers. I'm wondering if it has to do with that. Ill try using just one card for now. If that is it. I might just buy the 3870X2 .. i mean its crossfire on one board but not two boards. Also.. i dont see any chip in my board.. like SLI has a chip on some boards you have to reverse if you want SLI mode. I play other games like Crysis and such .. or i would not have a rig like this.

[Biz]

This site is great: http://www.playtool.com/pages/artifacts/artifacts.html

[Ughmahedhurtz]

Before you bother with swapping hardware around, honestly, I would run (do not walk) and grab Prime95 ('http://www.mersenneforum.org/showthread.php?t=6572'), CPU-Z ('http://www.cpuid.com/cpuz.php') and Core Temp ('http://www.alcpu.com/CoreTemp/') and verify the following:

1) CPU-Z reports that the memory timings and voltages are correct PER SPEC from your RAM manufacturer. If you bought no-name stuff that doesn't have specs listed, you're in for some interesting times.

2) Launch Core Temp and make sure it's reporting reasonable values. Once it's running, run Prime95 and let it run the Torture Test on both cores. Keep an eye on your core temps for the first half hour or so. If it goes above 70C, you're getting into dangerous territory. Sure, they're rated for 100C technically, but running them over 70 for any extended time is going to shorten the life of the CPU significantly.

3) Let Prime95 run overnight. When you get up in the morning, it should still be running. If it reported ANY errors, you have a hardware error probably NOT related to your graphics subsystem. At this point, contact your MB manufacturer and make sure the CPU, RAM, FSB and other frequency, timings and voltage settings are all correct for your setup.

If you can get through all three above without errors or overheating, then I'd start looking at graphics drivers or hardware as the culprit. The rationale is that memory corruption caused by bad RAM, a bad mobo or bad CPU can cause graphics artifacts and spontaneous reboots just like what you'd see with a bad video solution. Eliminate the easy stuff that just takes some time letting some software run before you go cutting your knuckles on heatsinks and shit.

[Chorizotarian]

Also you should be able to look in the event log and/or the feedback and Support control panel to see what crashed. From the symptoms it sounds like a graphics driver.

[turbopinto]

Today I went ahead and tried the method to reduce WoW latency, and ended up having several hard lockups just after, and during login. It only stopped doing the lockups after I removed what I had changed. I don't know if your problem is the same, but if someone else might be able to explain the problem I have, I'd appreciate it.

[Kissell13]

Sounds to me like a heat issue on your gpu. Check out your gpu temps as you play. I had the same issue and was able to solve it by taking off the top plastic part of the gpu heatsink and it was full of dust and not allowing the fan to spin at correct speeds and choking the air way up.

[Lyonheart]

Update: I had not read some of these suggestions for a few days. I removed one video card and still crashed. I tried a few dif versions of my drivers.. still crashed. The last thing i tried was adding more cooling ( 6 more fans ( i have cooler master stacker 830.. so i had lots of room for fans. Well that "seemed" to fix it.. i went several hours with zero crashed. Until just a few minutes ago.. so I'm thinking that the extra cooling helped narrow it down to some sort of healing issue. So thats where i will focus now..cpu/gpu/psu heating issues and ram as suggested.

[Lyonheart]

ok im not sure if this is right.. but your saying 70c no higher? right ow im running a test.. it shows all 4 CPUs and they are running 73-90ish .. ( was 65-72 while idol ) as soon as i started the test. this thing has water cooling.. is somthing wrong here ? hehe

Not sure if i said this.. but i bought this PC from a friend who built it for his video production company ( to do some editing at home ) he did the install.. and is no expert. He OCd the cpu and everything. Can i un-OC the cpu? I mean I'm not sure i need it to be OCd anyways.. if thats part of the problem

Update: i tested my other PC and its two cores run around 40c ish .. so I'm sure my super hot CPU is the culprit.. or it seems that way for now. I'm wondering if the water cooler is just not properly installed ? i mean I have no prob just yanking it out and getting a good air cooling solution.

[Dawnstone]

This may not be the same issue but...

My roomate just got a new PC and it was powering off randomly (in Windows, during boot up, while idling, or in a game)... we checked CPU, RAM, Hard Drive, OS, blah blah... today we swapped power supplies with another machine, and the other machine got the power-offs instead.

So if you got an extra power supply around, swap em and see if you can either A) not get the error anymore or B) get the same problem on another machine via the power supply swap.

[Lyonheart]

I think based on my testing its the CPU..( running way too hot .80-90c under load ) I think I got "punked" when I bought this from my so called friend lol.. I priced it out before I paid for it and parts alone came to 5200 ( he has receipts he gave me for 5500ish in Jan when he ordered them ) but today its still 5kish in parts alone. I paid 3200 for it. But I'm hoping the CPU isn't Fd.. that will bring it down to under 4 alone and it will seem like less of a "steal" then lol

[Lyonheart]

After running CPU tests for heat i concluded the thing was WAY to hot. The cores were averaging 80-90c and peaking at 99c under load 70-75 idol! I simply removed the water cooler block.. re applied new thermal paste.. snugged it all up and now my idol is 45-49c and under heavy load I'm averaging 60c peaks at 69 for a sec and stays floating between 58 and 61c! that is good right ? A lot better than it was! The guy who installed the system just seated the block poorly I guess.. I'm glad it was that simple.

I have not played yet.. so ill do that now and see if i can NOT crash today. ill report back those result. Thx for your tips and advice!

[Lyonheart]

No crashes in two days ? Im getting average of 72c per core when 5boxing though.. i need to get it lower.. but thats stable enough to not crash 8)

[Ughmahedhurtz]

After running CPU tests for heat i concluded the thing was WAY to hot. The cores were averaging 80-90c and peaking at 99c under load 70-75 idol! I simply removed the water cooler block.. re applied new thermal paste.. snugged it all up and now my idol is 45-49c and under heavy load I'm averaging 60c peaks at 69 for a sec and stays floating between 58 and 61c! that is good right ? A lot better than it was! The guy who installed the system just seated the block poorly I guess.. I'm glad it was that simple.

I have not played yet.. so ill do that now and see if i can NOT crash today. ill report back those result. Thx for your tips and advice!Excellent! Yes, 70-75C for short periods isn't a big deal but average load temps of 85+ is trouble. :)

Glad it was a simple fix.