P4 restarting itself, is it too hot?

[pepsimanz]

Running a P4 1.6a @ 2.4 with 150FSB and 1.75v setting (actual 1.68-1.7v). I have been running a lot of benchmarking and burn-in programs to make sure that the system is stable.

Everything is fine with Prime95 (running all day with no problem), Seti@home, Sandra, 3dmark, etc.

I tried running 3dmark looping demo with prime 95 running at the same time, no problem for the whole night. However, when I do 3dmark benchmark looping with prime 95, I noticed that my computer automatically restarted overnight when I wake up. But the everything will still run again and there is no problem. I also tried CPU burn as I heard it generates a lot of heat, but it seems stable. (although, I only used "normal" setting as the higher setting makes my computer really really slow)

I am suspecting this to be a temp problem because I know that the P4 are designed to throttle down and shut down the cpu when it's overheating. Unfortunately I got the MSI 645 ultra MB so I have no idea how hot my cpu (design defect). I am using stock heatsink and fan, with 1 intake and 1 exhaust case fan, and my case temp is around 34-36 idle 36-38 under load. In addittion, I have my Crucial DDR PC2100 running at ~ 225mhz (4:3 ram setting to make sure that it's not a ram problem).


So the question is, with my speed and voltage, should I get a better heatsink+ fan, is heat the reason why my computer restarts by itself? and if I get one, I am thinking about the avc sunflower and the new thermaltake 7+ . I don't want to get the new alpha because I don't like taking my MB out to install it. which one would you recommend? thank you in advance

 

[KenAF]

peps,

I'm sorry to be the one to tell you this...but you bought a lemon.

A vast number of people have reported the reboot issue with the MSI (and Abit BD7) on this forum and Hardforum.com. On Hardforum.com, there are probably 30 reports of just the issue you describe. There are plenty of reports here as well. The problem has nothing to do with heat or anything of the like; it's just a bug with the board's bios that results when overclocking. There is no ETA on when a fix might be available, if ever. Your best bet is probably to return the board, as others have done. Get yourself a P4B266-C, P4B266, or P4S333 instead.

And as a general FYI, unless you are running 1.9V or above, the "Northwood" P4 (1.6A/1.8A/2.0A/2.2A) isn't going to come close to overheating with the stock heatsink. It doesn't throttle til >60C and it doesn't reboot/shutdown until closer to 80C.

 

[Duvie]

OK...First I would like to say KenAF, Please reread!!!

he isn't having a reboot issue but a restart issue...Most sis boards also haven't been having them and not i the frequency the i845d mobos have....

I think you misinterpreted the post....


pepsimanz,

1)I do not think heat is an issue as 1) many mobos have reporting errors and my msi board told me in prime95 I was heating 60+c and having no issues ran it steady for 5 days....2)I thik those case/ system temps are obvious in error unless you are in a room of over 85f since those temps are closer to 97f...MSI has had reporting temp issues all over the place...Which bios???

2)If it makes it through prime and seti and 3dmark by itself I think it is fine...My take on this is try something else either then 3dmark...IN past I have always had mixed luck when ocing (at any level little or a lot) with 3dmark...3dmark2000 would give me an error at speed while 3dmark2k1 would run fine and even at higher oc...I don't trust that buggy software to begin...

3) I am not sure what you mean by "normal settings"...are you talking about memory settings???

 

[Z24]

What's your operating system? Win2k and WinXP have a "feature" turned on by default that automatically reboots after a blue screen of death.

To shut this off, go to System Properties-->Advanced-->Startup and Recovery Settings and uncheck Automatically Restart. (that's for WinXP)

Now, if you get a BSOD, you'll know for sure.

That may not be the problem, but it's something to explore.

 

[pepsimanz]

Even though I do have a restart problem, not a "reboot" problem, I still want to thank KenAF for his input because he made me decide to exchange my MSI 645 ultra for the ASUS p4S333. I like the Sis645 chipset, but the MSI 645 just seems problematic from the other forums (cpu temp prob, no mem voltage treak...), so if I can exchange (without restocking fee), I will do so and upgrade. I already sent in my RMA, hopefully, it will be approved.

In the mean time, I will still mess with this board a little bit more. and I guess from both KenAF and Duvie's input, it's not a heat problem as the P4 should be capable to run at 60C. But I was so tempted, I already jumped into the $19 sunflower deal after I posted this thread. Well, I guess it's never bad to have more than enough cooling.

Duvie:

1)I am using the version 1.2 bios and I will upgrade to the newest bios if I keep the board, however, I heard that the temp problem is still not fix with the newest version of BIOs. In general, I guess I shouldn't trust any of the temp reading.

2)I guess I will try a few more programs, but in general, I am happy with the stability. However, the restart problem that I am experiencing bothers me a bit, because if it's not heat, then I have no idea why it happened, and I surely don't want it to happen too often in the future.

3) I meant Normal process priority in CPUburn, it runs really really slow if I use high process priority, otherwise, it is stable even running CPUburn all night. Do you or anyone else have any more recommendation on what benchmark or burnin program that I can try?

Thank you both for your input.

 

[pepsimanz]

Thanks Z24,

I will try turn that setting off and run the same thing tonight and see if I see the blue screen tomorrow morning. I guess that way I will see if it's still a restart problem or just instability. Thanks again.

 

[MadRat]

After hearing so many cases about this spontaneous reboot problem, I think its the chipset causing the problem and NOT the processor.

 

[Amused]

Explain the difference between "reboot" and "restart."

If my computer restarts, it reboots.

 

[ogn]

You wouldn't happen to be using XP, PCanywhere, and/or Radeon combo are you?

 

[pepsimanz]

AmusedOne:

A "reboot" problem is when the computer does not boot up again properly after you tell it to reboot/restart in Windows.

My "restart" problem is when the computer reboot/restart itself without my instruction.

Ogn:

Yes, I am using Windows XP, but I am not using PCAnywhere or radeon vid card.

 

[MilkPowderR]

Reboot, Restart, Reboot, Restart, bluh bluh BLUH..!!!

Formula: Auto Reboot =(it means equal to) Auto Restart, Period.

I think what KenAF meant was Auto Reboot. He probably have said it fast and might've made a mistake in typing. We are assuming "Re"-booting by itself. Ken knew that and we should know that.

 

[Sid03]

does it only do it at 2400mhz? what if you set it back to default speeds?

you might just be pushing the system too hard. it's very possible for a system to run one app just fine, but choke on another.

 

[GFORCE100]

I would notch the Vcore up some more. you might just smile again.

 

[barryng]

It is my opinion the spontaneous reboot phenomena is related to both Windows XP and overclocking. The following has been my experience.

First, I did not experience this phenomena with Windows 98SE. It only became apparent when I "upgraded" to XP. I also noticed this problem with both my older Abit ST6/1.2Gig Tualatin and my current Abit TH7-II/1.8Gig Northwood.

Second, I have noticed that the frequency at which auto-rebooting occurs is related to how far I overclock my processor. I am currently running a 1.8Gig Northwood on a TH7-II. At rated speed and voltage it never auto reboots. As I increase the FSB frequency the tendency to auto-reboot becomes greater. I can mitigate this tendency, to some extent, by increasing the CPU voltage. I found I can run at 117Mhz FSB (2.1 Gig) and 1.6 volts without a problem. At 120 MHz FSB (2.16 Gig) it will auto-reboot occassionally even at 1.625 volts. At speeds faster than 120 MHz FSB it will even auto-reboot while Windows is loading or immediately thereafter.

Since I did not have this problem with 98SE it is easy to come to the conclusion that XP is prone to this problem. Also, since the extent of overclocking the CPU has a very repeatable effect, this also seems to be a contributing factor.

 

[GFORCE100]

Well it seems to be the board then, Windows XP is just software, it does what its told to and it the electricity flowing through the system can't be sustained 100% of the time within limits it will reboot. I've heard this about Abit boards quite a lot. Get an Epox 4BDA2+ We have one now and our P4 goes upto at least 2.75GHz stable in Windows 2000 and XP. We will be using it for the GF4 benchmarking.

 

[WHipLAsh13]

I have a 1.6a running at a 145 fsb and have never had one reboot/restart issue. I am using an Asus P4B266 mb. This system is rock stable. Not one issue.

 

[sandorski]

Bad/poor drivers can cause reboots. The newness of XP would likely have some driver problems, so though things are unstable now, in time and new driver releases stability may increase.