NEw PC reboot issues. Any ideas?

[lens pirate]

Mysterious reboot Please help
PROBLEM: after a few minuets of gaming (http://www.hardforum.com/showthread.php?p=1033613029#) ( left 4 Dead , Far cry2, Crysis) the computer will just black screen and reboot. When I run the Windows Experience Index updater it will also black screen and reboot. At all other times the system is ROCK SOLID stable.

I can run prime95 for hours and the CPU nevers got above 58 C.
System specs:

System O/S Vista 64
CPU i7 920 Not over clocked, Cool master V8 cooler
Ram Dominiator DDR3 1600 6 gig Running at STOCK speeds not over clocked to 1600
Motherboard Asus Rampage extreme2 805 Bios -Voltages set to auto
Power supply Ultra X3 1000 watt
Video 2x EVGA 280 lastest driver, not over clocked
XFI sound card that came with the motherboard
Case is Lian Li A70 with 6 big case fans

EVGA utility show idle temps for sound card as follows depending on room temp

slot one 45-50 C
slot two 37-42 C

CPU according to the Asus utility Idles right around 28-32 C depnding on room temp.

Currently at idle: CPU is 28C, South Bridge shows 44 C, North bride shows 47 C, top video card shows 48 C bottom card Shows 39 C Fan is at 60 percent.

I think I have a bad power supply or bad motherboard. Any suggestions?

[Faolan]

It sounds like it's your gfx card overheating. Quick way to test for this is to get a stress test application to check it out.

If it isn't then it's probably a PSU fault not being able to deliver enough juice.

[tim]

Get the free hardware monitor from http://www.cpuid.com . Run a 3D benchmarking program and watch the temps on the graphics card and the CPU. If one spikes that's probably the issue. Also get the free version memtestx86 (google it there's two main sites but don't pay) and run it overnight.

[lens pirate]

Get the free hardware monitor from http://www.cpuid.com . Run a 3D benchmarking program and watch the temps on the graphics card and the CPU. If one spikes that's probably the issue. Also get the free version memtestx86 (google it there's two main sites but don't pay) and run it overnight.


Will do

[OdiN1701]

I would say one of three things:

Video card issue - overheating or hardware problem with it.
RAM hardware problem - causes random reboots, more often when stressed.
Driver problem - update to all the latest drivers for video/chipset, etc.

[Zepher]

remove one of the video cards.

[lens pirate]

I would say one of three things:

Video card issue - overheating or hardware problem with it.
RAM hardware problem - causes random reboots, more often when stressed.
Driver problem - update to all the latest drivers for video/chipset, etc.


Its not the ram and CPU...tested the crap out of them.

Its either the video cards themselves or the powersupplies ability to feed the power hungry things. I am startinmg to thing its the video cards.

[WhyFi]

Your temps are fine, although I wouldn't be happy with 58 deg C at 100% load (CPU).

Never been a fan of Ultra PSUs. I would also vote for trying the cards one at a time.

[jetboy]

I wouldn't be happy about 58 deg either even though i think your cpu is a 130w. My 6000+ is 125w and overclocked and even at 29 deg C (85F) and 12hrs of dual primes running, I don't see over 55. I would think your cooler should be doing better than that. Either way, I don't think thats the issue. Are your video cards overclocked at all? What about RAM? Even a slight bump in either of these can cause this instability regardless of how hot its getting. My vid card does this even when its realtively cool (55C after max after hours of gaming). Bump the OC down about 8MHz. RAM is the same (depending on if I use CAS 5 or 4 even if underclocked). Pairs only multiply stability issues on both fronts. Test system with stock settings with one video card and 1 stick RAM. If it occures, test with different card and stick of RAM. Either way, you have to narrow it down to either a component, configuration, or a setting. This process can sometimes take a week to fully narrow down a issue and fully tweak the system. Luckily, I enjoy this portion of a system build. The wife on the other hand, lothes it. Good luck.

[DennisW1]

I can't speak directly for Vista because I don't have much experience with it, but Windows XP defaults to rebooting if the system encounters and error. It's a setting that can be toggled off in System/System Properties/Startup & Recovery/System Failure by unchecking the box that says "automatically restart"
Mind you I'm speaking of Windows XP, but I must assume there is something similar in Vista. This can drive you nuts as any serious system error, be it hardware or software, suddenly reboots your computer wihout warning. It is set to "on" (box is checked) by default.

[HankScorpio]

Are you sure you are running the latest forceware drivers 181.20? As anything previous to that would have issues with SLI on an X58 board. Do you get the same issues when running on a single card?

[Zepher]

Your temps are fine, although I wouldn't be happy with 58 deg C at 100% load (CPU).


I wouldn't be happy about 58 deg either even though i think your cpu is a 130w. My 6000+ is 125w and overclocked and even at 29 deg C (85F) and 12hrs of dual primes running, I don't see over 55. I would think your cooler should be doing better than that.


What are you using to monitor your CPU temps?
Real Temp (http://www.techpowerup.com/realtemp/) will show you the temp of each core.

Speed Fan (http://www.almico.com/speedfan.php)will show the temperature of the CPU, which is different than the core temps.

Using Speed Fan, I get a CPU temp of 25C at idle and 43C at 100% load.

Using Real Temp to monitor the cores, I get 40, 42, 40, 40 at idle and 63, 62, 60, 60 at 100% load using Prime95 (http://www.mersenne.org/freesoft/)

My Q6600 is overclocked slightly to 3.06Ghz from the stock 2.4ghz.
My cooler is an Arctic Cooling Freezer 7 Pro. (http://www.newegg.com/Product/Product.aspx?Item=N82E16835186134)

Ambient room temp is 23.1C

My eVGA GTX 260 SSC idles at 70C and when gaming hits 81C, which to me is kinda high.

[Bobster]

Grfx overheating will generally display artefacts rather than re-booting the machine..

[Moppie]

Grfx overheating will generally display artefacts rather than re-booting the machine..


I've had 2 different cards over heat and re-boot 2 different machines.

Neither displayed any artifacts, the machine would just crash mid game.

One was solved by adding extra cooling to the case, the other was found to be a manufacturing defect, and a recall had been issued for the card.

[Sean]

remove one of the video cards.

^^

Pull a card and test.

[lens pirate]

UPDATE:

Disabled SLI leaving both cards in the machine and it runs fine. Its looking more and more like a power supply limitation.

As a side note when I first Built this PC I had drives in it I was going to build a really big raid 5. But it the raid pack fail over and over again. Runs fine with one or two drives. Power again? I don't know and I have know why to test the PS at high load.

[OdiN1701]

1000W is plenty of power for your system.

Could be a power supply problem, but not due to wattage. Also could be SLI driver issues.

[lens pirate]

1000W is plenty of power for your system.

Could be a power supply problem, but not due to wattage. Also could be SLI driver issues.


Well as it turns out the problem came back with SLI disabled.

But Nvidia just released new drivers. 181.22 and they seem to have solved the issue. I just played left for dead with Eye candy maxed, resolution cranked way up and for almost a hour.... No trouble.

PC games and drivers are buggy as hell.

[BeritOlam]

I still wouldn't rule out your PS. The total watts output means nothing if the rails aren't stable. Ultra's are better than a cheap generic PS, but someone pushing your computer specs would typically be using a better-quality PS.

[lens pirate]

I still wouldn't rule out your PS. The total watts output means nothing if the rails aren't stable. Ultra's are better than a cheap generic PS, but someone pushing your computer specs would typically be using a better-quality PS.


Dam the reboots are back... Just tried to run the Windows Experience assesment tool and the machine reboot the instant it started the Aero Assesment.

[Moppie]

It really does sound like there is a hardware failure somewhere.

The powersupply would be a good suspect, as would one of the video cards, or ram.
But even a faulty bus controller on the board can give similar issues.


Fault testing can be time consuming and frustrating, but you need to do it if you are ever going to beat this problem.

Other members have outlined the basic steps to follow, and will get you on the right track. Once you have stripped the computer back to its basics, they will be able to help fully test everything :cool:

[Bobster]

the Ultra X3 has had good reviews, but still could be a faulty unit?

[BeritOlam]

Generally speaking, defective hardware will typically manifest itself in a frozen system. I've trouble-shot dozens of computers and talked to pros who deal with this kind of stuff on a regular basis, and they have confirmed this to be generally true.

On the other hand, a failing or defective PS will quite typically show up in the form of random reboots.

So playing the averages, the PS is probably the first thing you'd want to rule out in troubleshooting your problems. They make PS testers for around $10-20 that is the only surefire way to tell for sure. The newer PS's may require more advanced testers than the one I've used in the past.

[orion25]

I vote for the PSU being the problem. Get a PC Power and Cooling PSU since you are running a really powerful system. The silencer 750 (or is it 700) would be a good choice. Basically if you spent more on your HDD's or your CPU (within reason) than your PSU you will have problems.

[WhyFi]

PC P&C are great - I have a Silencer610, never had a problem, even with crossfire 3870s, 3 HDDs, quad-core CPU and 7 case fans... the Silencer 750 is on sale for about 100 bucks at NewEgg.

[orion25]

Never had a problem with my Silencer 750 as well. But PCP&C is not the only good PSU retailer out there.

Go to ocforums.com and look for their power supply reviews (may be combined with case reviews). If you have a bad PSU you have a bad system.

[BeritOlam]

If you have a bad PSU you have a bad system.

Sounds like someone speaking from experience! ;) ;) ;)

This is one area where people will often 'get skimpy' when building a new computer. After all, don't 500 watts = 500 watts, no matter what the PS? Well...no, they don't!!!

Now that these new(er) systems can easily push 700+ watts, the problems with a cheap PS only become more pronounced.

[ChasP505]

From experience... my bet is it's the PSU. Replace it with an Antec.

[lens pirate]

Well I agree. I ordered a toughpower 1200 based on reviews read over on HARDOC

I will let you guys know how it turns out. I will be retruning the Ultra x3 to ultra if this fixes my issues. I should end up with a new one back.

[OdiN1701]

Generally speaking, defective hardware will typically manifest itself in a frozen system. I've trouble-shot dozens of computers and talked to pros who deal with this kind of stuff on a regular basis, and they have confirmed this to be generally true.

On the other hand, a failing or defective PS will quite typically show up in the form of random reboots.

So playing the averages, the PS is probably the first thing you'd want to rule out in troubleshooting your problems. They make PS testers for around $10-20 that is the only surefire way to tell for sure. The newer PS's may require more advanced testers than the one I've used in the past.

Those PSU "tester" don't really do much. They do provide a load on the supply, but the only other thing they do is act as a jumper for the Power-On lead. That's it. I can do that with a paperclip to test if the PSU will power on. That's all they really test is if the PSU powers on when the pins are jumped.

Did you run Memtest on your RAM? Do that if you haven't.

I still would suspect something with video cards or drivers though - it seems that's only when you have a problem is when it's doing something with the video system.

[lens pirate]

Thanks guys it was the PSU. replaced it and all is well!

[ChasP505]

Good news!

[BeritOlam]

Glad you solved your problem!