System rebooting

[stevesmithx]

My system keeps on rebooting after some time booting into windows(2 hours,sometimes 1 hour) .
I know it is not a virus for sure as i have multiple OSes and the problem
occurs in all OSes.So it must be a hardware problem.Can anyone say what problem in the h/w generally causes this problem.
Please help...

[vart]

heat? Motherboard could decide that the CPU is overheatning... and reboot

Powersupply problem - some are too smart...

[stevesmithx]

Thanks for the quick reply vart.
I seriously doubt that this could be the problem because
the system restarts whenever i run processes consuming
high resources.
What can i do to solve this problem?
Which part of h/w should I check?

[vart]

the system restarts whenever i run processes consuming
high resources.

It makes me beleive I'm right

What can i do to solve this problem?
Start with checking the coolers - maybe your CPU cooler needs to be replaced or oiled...
Maybe your whole system is missing additional cooler, or some cooler is out of order or incorrecly set

[stevesmithx]

Thanks again vart.
By "Cooler" do you mean the CPU fan?

[PING]

By "Cooler" do you mean the CPU fan?

Yes, i think he does. You might want to check the power supply as well as the socket where you plug in your computer. The same used to happen with me and it turned out that the socket where i used to plug in the computer was bad and the voltage used to fluctuate a bit causing the system to reboot. Also, check your cpu temperature at high load. If you are using windows, you can get a temperature monitor like speedfan. If you are on Linux, you could use something like GKrellM.

[dwks]

Also check your output vent, wherever the air that your fan blows comes out . . . my laptop gets very hot very quickly if I accidentally block it. Just having, say, half an inch of space around the tower or whatever might make the difference.

[SlyMaelstrom]

By "Cooler" do you mean the CPU fan?Yes, you'd both be referring to the same thing, however "cooler" is somewhat more accurate than "fan" as the object that is cooling your CPU is actually a heatsink with a fan mounted on it. Generally the latter of the two is more optional (but on your standard PC there is a fan). This is generally either referred to as the HSF (Heatsink/Fan) or CPU cooler, if you prefer.

Now, onto your problem... Vart is pretty much on the right track. The reboot would come down to either one of your main components overheating or your power supply being overloaded.

Power supplies wear out in time and their original wattage capacity will decrease dramatically if you put too much pressure on it (let's say pressure is a measure of (time * PC Usage)). Typically this is not a concern for a purchaser of a manufactured PC as they will give you a power supply great enough that it should last you several years of use before you notice wear. However, if at any point you might have added a significant amount of RAM, or a video card, or a sound card, or an additional hard drive... you would increase the power consumption and increase the wear of the power supply.

Heating issues that would cause reboot would generally come down to either your CPU, North Bridge, South Bridge, or RAM. Typically, you can rule out the latter two options as they never draw that much heat unless you are severely overclocking. Even then, the south bridge is generally very cool. The CPU and North Bridge, however, on modern computer require a significant amount of cooling. If at any point the PC was jarred enough that the heatsink on either or both of these components came lose (even slightly) then the reduced contact between the heatsink and the chip could cause it to overheat and shut down the computer. It's also possible that through time the thermal paste used between the heatsink and component has worn out and needs to be reapplied.

I would first follow Vart's advice and get a temperature monitor that will tell you the temps of all your main components. Test them both idle and under load. Then you should search the internet for the proper temperatures for your particular CPU. If you are in the danger area, then you know it's a heating issue. If your temperatures seem fine, then it would likely be your power supply... however, before you go out and buy a new power supply... make sure you know what you're buying and how to install it. Also, make absolutely sure this isn't a software issue before you spend a dollar.

[stevesmithx]

First of all,
Thank you all.
I installed speedfan s/w which shows the remote(i don't know what it is) to be
in a high temperature(52C) when the System is in idle state.However the remote becomes 70C
when it is in loaded state and the system restarts.[I opened 3 instances of Turbo C to load the system ;) ->showed CPU usage-100%].I found it rather strange that my CPU fan which used to make much noise every time I open Turbo C before didn't make much noise this time.
I had a problem with my UPS 2 days ago(backup time went low from 15 minutes to 10 seconds) and I replaced the battery alone for the UPS.
So I am in a dilemma whether the problem is with the cooler or the UPS.
If anyone have any suggestions I would very much appreciate it.
Thanks in advance.

[Bubba]

Moved to Tech board.

[SlyMaelstrom]

I installed speedfan s/w which shows the remote(i don't know what it is) to be
in a high temperature(52C) when the System is in idle state. However the remote becomes 70CThis doesn't tell us too much unless you tell us what your CPU is... more of the C2D line would say 70C is very high... however, on the older P4s, specifically the Prescotts, that isn't such a bad temp. The AMD CPUs can vary in that range, too.

[brewbuck]

So I am in a dilemma whether the problem is with the cooler or the UPS.
If anyone have any suggestions I would very much appreciate it.
Thanks in advance.

Well, you can prove the UPS theory by plugging straight into the wall and see if it stops the random reboots.

If not, and CPU fans seem okay, air flow seems okay, then suspect the power supply. If you have another sitting around, swap it out and see what happens.

[PING]

So I am in a dilemma whether the problem is with the cooler or the UPS.

I had the same problem but with the spike-guard, so i plugged the computer directly into the wall socket and the reboots stopped. Your temperatures are a little on the higher side if you have an original Intel. Also, running TurboC doesn't put your system on full load, it just appears to be doing so. You should try running a resource hogging game or maybe convert a few movies from one format to the other to really test your system.

[matsp]

It is unlikely that any part of the system decides to REBOOT on a "too hot". It may be that the processor gets too hot and triple-faults. Triple fault is when the processor sees a fault (page-fault for example), and in the process getting to the fault-handler, the processor encounters another fault-condition (we're now on the road to the double-fault handler), and when dealing with the double-fault encounters another fault. Generally, this happens when you have completely messed up some important data structure(s) inside the kernel space, such as the interrupt descriptor table, global descriptor table or the kernel-side page-tables.

One of the ways to unintentionally cause a triple-fault is to have the processor get too hot and write back bad data to the memory. Or by having memory that slowly looses it's content. Most of the time, both of these will lead to blue-screens/kernel oopses before leading to triple-faults, but sometimes the processor manages to hit the same "perfect" spot every time.

All motherboards will "deal" with a triple-fault by rebooting the machine [in fact, the shortest code to reboot the kernel is something like this:

.data
nulidt:
word 0
dword 0

.code
lidt nulidt
int 3

(For those not particularly familiar with x86 assembler, the above code sets the interrupt descriptor table to be zero lenght, then calls interrupt number 3 - which of course doesn't exist in a zero length table, so the processor tries a double fault, which tries IDT entry number 8, which also doesn't exist, so processor gives up and issues a special bus cycle indicating "triple fault" - the chipset detects this special cycle and reboots the machine).

--
Mats

[maxorator]

It is unlikely that any part of the system decides to REBOOT on a "too hot".
It is not unlikely, since it has happened to me. Before I made a BIOS update my computer tended to reboot/shut down at high load and when I turned it back on, the vents were crazy (sounded like a vacuum cleaner) and BIOS showed that the CPU temperature was 120. Probably BIOS interpreted the information from indicators a bit wrong since the computer would've been on fire if the temperature would've been 120. And after a BIOS update everything was OK.

[Elysia]

By default, Windows reboots whenever there's a blue screen (which is usually stupid). You can try to check if you get one by undoing this default behavior.
Go into system, advanced tab -> start (and restore)?, and untick Reboot Automatically.

[stevesmithx]

Thank u all .
Yesterday the problem got worse and system simply went off.
I tried to switch on the system but to no avail.It simply won't switch on.
But today when i switched it on it budged.But I am afraid my system would reboot again
at any time even as i type this.
Go into system, advanced tab -> start (and restore)?, and untick Reboot Automatically.
I did that and waiting for the blue screen..

[matsp]

Thank u all .
Yesterday the problem got worse and system simply went off.
I tried to switch on the system but to no avail.It simply won't switch on.
But today when i switched it on it budged.But I am afraid my system would reboot again
at any time even as i type this.

I did that and waiting for the blue screen..

Sounds like a broken power supply - but could be any number of other components related to getting the system to run at all.

--
Mats

[stevesmithx]

Thanks all.
The problem was in SMPS.
I bought a new one and my sys is working pretty well.
Thanks again all for your time and help.