Over the last few months I have picked up on some info that has me reconcidering the value of overclocking. Most of us by now already know that overclocking is risky to begin with, but that it could be done safely if you use a little caution and restraint. I am rethinking this these days and here is why.

This morning I read a forum discussion on P4 failures at Overclocker's .Com and I remembered reading about an increase in the number of motherboards that have been failing do to capacitor failure reported a few months ago at the same site.

The author of the news about increased mobo cap failure suspected that it was an engineering issue related to higher FSB frequencies because capacitor failure was virtually unheard of prior to the development of the100mhz FSB.

Furthermore, I have noticed a much wider interest in overclocking from our readers and I know that it is becoming a standard "thing to do".

What I am starting to believe is that motherboard manufacturers are starting to reduce the level of "over-engineering" at the component level in order to keep costs down, and that the rapid pace of development and the evolution of Industry Standards was made this a neccessity for the manufacturers in order to stay in business. They simply have no money to waste so they are cutting it very close when it comes to the tolerance levels of the components.

I have said this before as have many other guys on this site, "when you overclock your system via FSB and Vcore adjustments, you are overclocking the ENTIRE system. Entire means everything.

So, if you agree to what I just said, then think about this forum discussion about P4 failures and imagine that it could be the motherboard components that are failing first, capacitors and mophets, which regulate proper power usage to all the other components. They weaken, and begin to kill others.

The guy with the USB issue, ok, the power was killing his USB controller module of the SouthBridge chip.

The point is, that it isn't just the CPUs that are at risk, these guys are reporting motherboard, RAM, and other componant damage as well.

My system is currently overclocked.
I am not telling you guys not to overclock.
I AM saying that there is more at risk when you OC then just your CPU and I myself am beginning to lean toward the "it ain't worth it" attitude.
My system is pretty fast a stock speeds. That money could pay for a lot of modding

you are right abou that lcpiper, my brother has been getting some boards back with unstable power complaints and i told him that his customers were oc'in them. he didnt belive me and said that it was the powersupplies that this customer was useing. i talked to him and found out that he was indeed oc'in his fsb and core voltage. then told him that we couldn't refund him his money and that we would send the boards back to him at his charge. man did he lose it, he was yelling at the top of his lungs at me.

i just said sorry bro i guess you didn't read our return policy " no refunds or exchanges with out a reciept or after 30 days, you will have to contact the mb manufacture ".

usually we would let it slide but he was saying that we sell bad product and that it was our fault... no way all of our product is top notch...

oc if you want to brag about it but in the end you will all ways lose something, and it will cost you. maybe a new system, maybe just a few parts, but it will cost you.

Hey Piper -

if we are going to quit overclocking .......... we better start making them damn quiet!

I won't stop water cooling, but I have to have a tangible reason for doing so!

Would have been nice if those guys had given there cooling methods (is it a heat problem).

Also, most of those rigs were volt modded? Are people seeing failures within the voltage range allotted by the mobo via the bios?

One casualty I have suffered due to overclocking was a GeForce3. It ran great for a while... I had the FSB set really high on the board PLUS I was oc'ing the card itself. The end result was a card that was very unstable. I wasn't able to run it in a OC'ed system at all after that. A friend of mine can still use it in a standard system running a meager 100mhz fsb (much different then the previous 187fsb home it was in before) at default speeds but once you try to put it in a faster system then it starts to fail again. As a result I now am much more conservative about overclocking video cards. i will probably be removing the waterblock for my vidcard from my h2o cooling setup as well.

Putwig, you guys have the same info I have from that forum link.

Most of them don't really give enough details, but yes, voltage looks to be the biggest factor. I am thinking that we in general have been a little too cavalier when it comes to raising our voltages for increased stability.

It also makes a lot of sense, why increase your voltage by sometimes 1.0V or more, when a PSU has to be able to produce clean power within much tighter tolarances.

Lol, I can't wait. Pretty soon we will have a thread where we are all bragging and reporting out system noise levels " in the middle of prime and UNDER 20db ! yaCan't HeaR That ! LOL "

Stop OCing? Never. But I have learned that I don't need to run the system on edge all the time. I always find out what the max a rig will take, then back it off quite a bit, usually 50% - 70% of the difference between stock and max. That includes my video cards. I've found a niche there where I get increased performance, system stability, and my temps aren't out of control.

And yes, I am very stingy when it comes to voltages. The only component I may juice too much is my RAM. I may be shortening the life of the stick, but it sure does increase system stability.

I have got my voltage maxed, or rather maxed at the bios level.

I have never "voltage modded" my board.

So, I have been running at 1.85 vcore for at least 6 months
(MBM5 reports 1.92 vcore). Now that it is a $100 chip I will probably leave it there.

This was the first rig I ever oc'd (also the first I built, the first I modded, and the first I water cooled .......... nothing like jumping in with both feet!).

I don't know if my next rig will have the juice turned up, I expect it will (not like we did not know there were risks!).

I guess I will wait and see where this all goes!

I will never stop overclocking, OC4life Yo!

I hear what you are saying though and I dont OC my 4 SMP rigs very much or my laptop, they dont OC well anyways, I have 2 rigs though that i DO OC on and play with different tweaking strategies.

(right now its a pair of athlon 1600xp rigs, 1 is an sis chipset and the other is a via, hopefully soon i'll be adding a P4 to my tweaking arsenal)

if you can afford 2 rigs, tweak one and keep the other @stable stock settings

if you can afford only 1 rig, if you OC it you are asking for trouble

just my .02

Or in my case, if you can't afford any and have a bunch, tweak them all and damn the torpedoes!

There is also a little difference between these dying P4s and the AMD chips that most of us run with.

The default voltage for most of our CPUs is 1.75V, but those P4s usually run around 1.6V or even less. And of course most of our current CPUs are .19 micron, not .13s. That does tend to make a difference I would think.

I still have to laugh though for that fool who had his system too high, killed his USB controller, moved his mouse to his PS2 port, and thought that he had cured his problem, lol

That was funny!

He actually thought the whole problem might be due to the prevalence of USB mice!

Read that thread through buddy.

This other said something so stupid that you have to assume that he was just BSing it or kicks.

He actually said that "burning in" a new CPU would lower the die size of the core, like from .19micron to .17

So go figure

Any n00b knows that heat expands, so his core would go from .19 to around .21. It needs to be wet in order for it to shrink. And you wouldn't be able to "burn-in" a CPU if it was wet because the fire would go out.