CPU temps, Nic cards, and other misc junk

Avder · Post by **Avder** » Tue Sep 07, 2004 7:26 pm

Question one: Does anyone know a good free(non nagware, non expireware)utility for windows that can monitor the bios of a P3 based system and display the CPU/MOBO temps somewhere on the desktop/system tray?

Question two: Can a bad nic card, even when disabled in device manager, cause a computer to hard lock(freeze)?

Question three: At what speed does the AGP Bus generally become unstable? Does this affect the computer even if there is no AGP card present(PCI Video) ?

Question 4: Does anyone know a good, almost instant way of telling if a CPU is being overclocked too far? Like, Computational error sorta tests?

Tetrad · Post by **Tetrad** » Tue Sep 07, 2004 9:01 pm

Avder wrote:Question one: Does anyone know a good free(non nagware, non expireware)utility for windows that can monitor the bios of a P3 based system and display the CPU/MOBO temps somewhere on the desktop/system tray?

http://mbm.livewiredev.com/

Avder · Post by **Avder** » Tue Sep 07, 2004 9:11 pm

One down, three to go.

AceCombat · Post by **AceCombat** » Tue Sep 07, 2004 9:28 pm

Avder wrote:
Question two: Can a bad nic card, even when disabled in device manager, cause a computer to hard lock(freeze)?

this depends on whether the card is truly "Disabled" just because it says "Disabled" in a operating system, does not mean it is physcially disabled and turned off.

Avder wrote:Question three: At what speed does the AGP Bus generally become unstable? Does this affect the computer even if there is no AGP card present(PCI Video) ?

this depends on whether the PCI Card is bridged through the AGP Chipset or not. <-- best guess answer!

Avder wrote:Question 4: Does anyone know a good, almost instant way of telling if a CPU is being overclocked too far? Like, Computational error sorta tests?

not that i know of

Mobius · Post by **Mobius** » Tue Sep 07, 2004 10:19 pm

The only instant way to know if it's OCed too much is to overclock it until it doesn't boot!

Long runs of Prime95 will establish is a CPU is stable at the speed you chose, but there's no way of knowing how much is too much, unless the system crashes.

If you can run Prime95 for 12 straight hours, then that should be stable enough.

The AGP bus won't become unstable with overclocking, but your AGP card might become unstable. The bus is simply the transport mechanism, and scales with the FSB multiplier usually. (Certainly on a P3 system it will).

PCI is normally 33 MHz and AGP is 66MHz. You can calc what values your PCI and AGP buses will run at by the diver ratios. If your FSB is 133, then your PCI ratio is 1:4 while the AGP ratio is 1:2. So, if you push your FSB to 145, then your PCI bus will be at 36.25MHz while the AGP will be running at 72.5MHz.

Remember, this is overclocking your PCI and AGP cards too. Many PCI/AGP cards can take a significant overclock with no problems. I had by AGP clock running at 75MHz for bloody ages.

However... some cards just won't like to be run far from the standard PCI/AGP spec. A good example is a TV-Tuner card. In my experience, they will only handle a very small overclock before crashing the system. Most sound and video boards overclock just fine, out to 37.5/75MHz, but I wouldn't expect you'd get any more than that no matter what cards you have installed. (Of course, adding more cooling, heatsinks, and possibly doing a volt-mod on the GFx might help you if it's the video card which is holding you back.

To find out what card is causing you trouble reaching a max-overclock, leave every card except GFx out of the loop (Take 'em out) and push for the sky. When you get it stable, add PCI cards in, one at a time, to see what the maximum is for each PCI card.

Avder · Post by **Avder** » Tue Sep 07, 2004 10:31 pm

Sys Info: P3, 133x7, PCI Bus is listed as 33 in the BIOS, while AGP is listed as running at 88.

Avder · Post by **Avder** » Wed Sep 08, 2004 8:46 am

Also, is temperature at all a decent indicator of how stable a CPU will be at its current speed?

fliptw · Post by **fliptw** » Wed Sep 08, 2004 11:00 am

some motherboards allow you to change the divider/multipler for the AGP and PCI buses.

some let you specifiy the exact MHZ each bus runs at independant of each other.

Avder · Post by **Avder** » Wed Sep 08, 2004 11:30 am

This one doesnt. But oddly, if I drop the bus down to 132, the pci shoots up like 15mhz. Its bizzare-o.

AceCombat · Post by **AceCombat** » Wed Sep 08, 2004 12:48 pm

Avder wrote:Also, is temperature at all a decent indicator of how stable a CPU will be at its current speed?

generaly Temps will just tell you if you have sufficeint cooling methods for overclocking. ive never heard of anyone basing their OC speeds on temps.

while higher temps can possibly lead to bad effects on a device and/or permanent damage to the device, its not used as a indicator on overall stability, in a more technical aspect it "CAN" show stability based on temp readouts and will alert you to high temp warnings and when to stop overclocking to prevent damage or enhance your cooling methods to keep the temps in operating ranges.

Avder · Post by **Avder** » Wed Sep 08, 2004 2:31 pm

Someone gonna confirm or deny what Ace is saying?

Avder · Post by **Avder** » Wed Sep 08, 2004 10:57 pm

Well the comp in question has been running stable now for over 24 hours...so I guess the NIC I pulled out was guilty of some kind of instability somewhere.

Vindicator · Post by **Vindicator** » Thu Sep 09, 2004 1:09 am

AceCombat wrote:
Avder wrote:Also, is temperature at all a decent indicator of how stable a CPU will be at its current speed?
generaly Temps will just tell you if you have sufficeint cooling methods for overclocking. ive never heard of anyone basing their OC speeds on temps.

while higher temps can possibly lead to bad effects on a device and/or permanent damage to the device, its not used as a indicator on overall stability, in a more technical aspect it "CAN" show stability based on temp readouts and will alert you to high temp warnings and when to stop overclocking to prevent damage or enhance your cooling methods to keep the temps in operating ranges.

I'll second that. Low CPU temps wont guarantee a stable overclock, but high CPU temps will definitely affect the stability of an overclock.

AceCombat · Post by **AceCombat** » Thu Sep 09, 2004 1:40 pm

Avder wrote:Well the comp in question has been running stable now for over 24 hours...so I guess the NIC I pulled out was guilty of some kind of instability somewhere.

i was hunching on the fact that you said the NIC was a likely cause, of your problems.

as i said before, just because its listed as "Disabled" does not mean it is physically turned off and not doing anything

suicide eddie · Post by **suicide eddie** » Thu Sep 09, 2004 2:14 pm

{Question 4: Does anyone know a good, almost instant way of telling if a CPU is being overclocked too far?}

smoke rings

AceCombat · Post by **AceCombat** » Thu Sep 09, 2004 5:15 pm

hahaha, yah thats a good way

Avder · Post by **Avder** » Thu Sep 09, 2004 11:35 pm

Well, two days so far its been up.

Best guess on if it would be safe to put an old voodoo3 in the AGP slot, given that AGP is currently forced to 88MHz?

Krom · Post by **Krom** » Fri Sep 10, 2004 5:21 am

My old voodoo3 3000 tollerated 83 MHz AGP with no problems.

AceCombat · Post by **AceCombat** » Fri Sep 10, 2004 12:05 pm

hey krom......got a question for ya located here:

phpbb/viewtopic.php?t=3526

is there a AGP/PCI-X adaptor. ive noticed alot of the new AGP Video cards say they have PCI-X Support.

what does that mean?

fliptw · Post by **fliptw** » Fri Sep 10, 2004 1:10 pm

some cards use a PCI-X to AGP bridge, they have the PCI-X connector, but they use AGP internally.

CPU temps, Nic cards, and other misc junk

CPU temps, Nic cards, and other misc junk

Re: CPU temps, Nic cards, and other misc junk

Re: CPU temps, Nic cards, and other misc junk