awesome computer and already have problems :)
sublime tool
hey guys , iv gotten my new pc , etc it runs fine...except for its random shut downs without warning...it usually happens when im playing battlefield 2... happened twice when i wasnt playing games. at first i thought it was a overheating problem, so i ran pcprobe 2 provided by asus and checked the temperature etc. the Motherboard is usually 45º Centigrade and the CPU is usually 33º Celsius. Some minutes ago i got a warning from the CPU fan saying it was having high rpm...
I really dont know what to do...this computer has dual SLI 7950 , 2 gigs of fastest ram and Pentium D 3.4 , i really dont know what is causing this shut downs
I really dont know what to do...this computer has dual SLI 7950 , 2 gigs of fastest ram and Pentium D 3.4 , i really dont know what is causing this shut downs
Replies
If there is an alarm you can set for overtemp, try that. The unfortunate part about auto-shutdowns is the lack of feedback. If you can set up a beep prior, then you'll know if it's heat or something else.
I'll also assume you haven't overclocked anything
Could be a powersource problem. That's a pretty powerful system, so what is your powersource?
[/ QUOTE ]
Yeah, thats what it sounds like. Try disconnecting everything unneeded that uses power (cd drives, extra harddrives, all sorts of external usb devices) and see if the problem persists.
Sometimes a PSU can not keep up with short extreme powerdrains, thus resulting in a complete shutdown.
If that isn't the problem, maybe check the RAM, but ram errors just like overheating problems usually result in a corrupted display or other kind of strange behaviour, and not that often in sudden reboots (but it can happen).
a while back my old psu started shuting down during windows.. and then it started shutting down right as the system powered up, i then reset the cmos which reset my speeds back to normal (was overclocking my cpu 700+ mhz and had my ram at 3.6 volts, it was ddr utt-bh5 voltage sucking ram). after that guess what? the system started up fine.
however after a while the psu died with a great big blue puff of smoke and it was time for a new one.
I've also had this happen with my athlon 64 x2, aparently the default cpu voltages weren't correct or something like that so u had to give it more juice so it'd run propperly... thats all i ended up doing and it was fine... so i dunno i doubt its ur cpu, most likely its the power supply.
Oh well, i can try to fix it myself, is it easy ?
P.S - thank you so much for the support guys !
The psu is usually connected by 4 screws in the back where the main fan is.
But have you tried disconnecting devices first? And what is the current PSU you have? Oh and it is much more important to get a quality brand with a good energy efficiency and a high "combined watt" power than just a high watt number (also because you will pay through your nose when the next powerbill comes
ok, the technitian informed me that it is 650 watts , and heavy ones , so i guess it rules out psu problems
[/ QUOTE ]Not necessarily, get yourself another PSU anyway to have as a back up (if you have the cash to spare).
Once you get it just connect it up and test the system, if it runs ok then you know the old PSU was faulty and probably best removed before it shags the hardware with it's shockstick.
Again, check the cooling fans... try running the computer with the side of the case off (If you aren't already). Or take it back to that shop and let him debug it
It may be the drivers as others have suggested. Usually that results in a blue screen/memory dump though. The nvidia site only has one driver for the newer cards. The link is http://www.nvidia.com/object/winxp_2k_91.47.html
The problem could relate to the dual sli. Will the computer work for a long time, then crash only when you try games? It could be that any time you use graphics acceleration, your computer bites it. As I referenced in the past, dual SLI is still a little rough. Unfortunately, I don't think it would be easy for you to pull out one card to see if the problem goes away. Most boards seem to have jumper settings and bridge cards.
If the problem is related to the RAM then choosing less demanding timings (and truning off dual channel) could fix the problem, or at least tell you where the problem is.
edit: if it really only happens with bf2 than i wouldn't rule out that buggy pice of shit as the source of the problem either
it could also quite possibly be your windows install, bf2 no longer worked for me at one point (gave me a bsod) but that was fixed with a windows repair.
PS! If you have a good case and take one side off the cooling will get worse because of incorrect airflow.
Edit: Cross post
Actually, I just found this: http://forums.nvidia.com/index.php?showtopic=18626 It may help you to try older drivers apparently
Update: Hmmm, actually I found another article that was very interesting... I haven't completely read it yet, but I thought I'd post it now: http://www.extremetech.com/article2/0%2C1558%2C1932947%2C00.asp
[ QUOTE ]
As we noted earlier, you can prevent these shutdowns by using a molex-to-PCI Express graphics card adapter and sharing the load with another 12V rail. Alternatively, just load up the other rails with more hard drives or other power-drawing peripherals.
[/ QUOTE ]
So, unless you read that article fully, it seems like you need a new power supply, when in fact you may not. Sounds like you just need to make sure the video cards don't share the same power cable from the supply. My guess is that you are right now since that typically makes the most sense when organizing the cables in a computer.
notman you have been a great help man ! so i guess yah..its the damn power supply, and the compouter technitian also considered it since on his tests he didnt push the graphic cards...agh...still keep em coming
/me picks up phone and dial technitian number.
No 'technician' should have to push a graphics card in a bench test. Graphics cards actually have a bios that does warm up tests during post. You would have been notified of any error. When hardware is in question there is very rarely a grey area, either the device works or it doesn't. If any of the major components do not work, the system will simply not try to run on faulty hardware, to protect the other devices.
Another thing you can do, is re-seat the cards. Meaning, remove them fully, and replace them securely. Also make sure the motherboard was installed on risers. Some do-it-yourself pc builders neglect this step, and it leads to shorts.
There is an extremely rare issue with motherboard drivers that control the standby or shutdown software functions as well. Update the mobo drivers if you have not.
Hope it helps.
Running two instances of Prime95 will will quickly tell you if your processor or ram is having probs. As someone else said, running 3dmark in loops should max out your gfx cards. If both of these work fine separately then run both the dual Prime95's as well as 3dMark looping at the same time, if it shuts down now its probably overheating something from the combined heat or the power supply is too weak. leave task manager open so you can view cpu's to make sure they are maxed out as well as nvidia control panel and a cpu temp prog to watch all temps.
There also are 'grey' areas in hardware. Filter capacitors can fail and cause the hardware to act flaky, but not completely fail. Many times random shutdowns can be simply temperature related. The motherboards are designed to shut down instantly if it detects an overheat of the CPU. This is done on purpose because CPUs can't handle overheating. I've seen this happen many times due to people having tumbleweeds in their heatsinks.
It does sound like overheating is out though. The PSU routing is probably the problem. From what I can tell, the SLI rated PSUs just do a better job of seperating the 12V rails. That is why I was suggesting he connect to individual lines from the PSU rather than one shared line. That should help seperate the rails.
Johny - Its always good to have as few things on any given rail (rail = power lead coming from PSU) as possible and if it IS an SLI PSU it should have 2 seperate rails marked specificaly for the GFX cards. If you can open up your case and have a look at the brand and model of the power supply then we may be able to give you more sound advice.
Again, good luck
ill keep you guys updated
Good luck.
The reccomended watts for the high end video card confs are way over rated because of some junk manufacturers over rating their shitty PSU's.
it may also be your chipset overheating, at one time i had a wire stuck in the chipset fan... and it led to all kinds of problems ranging from blue screen of death, freezes, random restarts, errors in programs...etc. thankfully i found it before any permenant damage was done to the chipset... but sadly my windows install was dead from so many errors.
be sure to check those things.
and by the way : thanks alot guys