Risks of Overclocking

[Caldenfor]

Since I have never overclocked, I decided I wanted to do some research on overclocking and I have been browsing the different threads on these forums. It may just be me, but would there be any way to get a sticky to the risks of overclocking? I myself have heard it can cut the life of your CPU in half, but well... I have never found out first hand =0

What exactly are the risks or dangers of overclocking?

Of course I will continue to do my research, but for those new to the forums it is something I know a bunch of people at least look into. I did read through some of the OC guide sticky, but I didn't find a nice list risks, but it was quite the guide to setup an overclock.

Thanks,
Dave

 

[Zenoth]

I guess that in most cases it'd be barely noticeable (life span decrease), but if you do extreme stuff, constantly, and let run your un-stable system as long as it can until it crashes, and start up again and retry extreme OC's, then I guess you can surely "break" something, fry something, or make it 100% non-responsive (kill it, simply said, from within, like the circuits for example).

I don't know anything about "Elite OC'ing" with Phase stuff and Liquid cooling and such, but I do know that I've always OC'ed all my CPU's and GPU's to date, to basic extents, and haven't had any issues whatsoever, and never saw any of my hardware break sooner than expected.

 

[myocardia]

Zenoth is exactly right. The higher you try to "push" your overclock, the more stress you put on every part of your system, specifically: the power supply, the motherboard, the processor, and the RAM. A 200-400 mhz overclock isn't going to shorten any of those parts' lifetimes enough to care about. An 800-1,200 Mhz overclock will.

 

[Caldenfor]

Okay sounds good, I am just trying to find out all the negative parts of overclocking before I attempt, if I attempt, to do it on my new E6400.

Dave

 

[Conky]

I would see how far you can overclock without raising voltages. This is the mark of the best overclocks to me and, in theory, if you aren't over-volting your cpu you shouldn't be damaging it. This is my theory anyway, lol.

Right now I am running a very modest overclock on my E6400 (300FSB for 2.4GHz) with no added voltage and it only runs a degree or two hotter. I don't see how it can hurt anything. When I get a better overclocking motherboard I intend to bring it to 400FSB, hopefully without extra voltage and a high-powered heatsink/fan on top of it.

 

[Atheus]

If you're not raising the voltages you won't damage anything. Even if you are raising voltage, it's pretty difficult to do actual permanent damage. Your motherboard is not designed to fry processors.

 

[F1shF4t]

Originally posted by: Atheus
If you're not raising the voltages you won't damage anything. Even if you are raising voltage, it's pretty difficult to do actual permanent damage. Your motherboard is not designed to fry processors.

If your not raising voltages, the extra clock speed will still damage the chip, neglidgible ammount as long as u keep it cool. Raising voltages will reduse the life of the chip even more. As long as the life of the chip is longer than its practical use it should be fine.

But with this its like everything, the safest thing to do is not turn it on at all, even operating at stock everything still degrades etc. That damage takes so long to accumilate that by then u prolly sold or chucked out the chip out anyways.

 

[LittleNemoNES]

this reminds me -- we need a Quick and dirty guide to C2D OCing.

 

[cmdrdredd]

Originally posted by: gersson
this reminds me -- we need a Quick and dirty guide to C2D OCing.

I was thinking about that. I'm not that good of a writer so I couldn't really write it the way I can do it.

However, basically it's about the same as the A64 systems.

You find your max FSB by lowering everything you can to the lowest settings (CPU multi and memory ratio at 1:1) and see when windows doesn't boot. Then you can find your max CPU speed by raising FSB with CPU on it's highest multi until you can't boot and generally aren't stable.

In reality it's easiest to set memory on 1:1 leave multi alone and start at near 3Ghz. So if you have an E6400 you can do 380FSB and default multi of 8x for 3040Mhz CPU speed. In general every CPU even a E6300 will get to 3Ghz by just putting memory at 1:1 and putting the FSb up on the speed that will put the CPU speed at the 3Ghz level.

I find C2D overclocking much easier than A64 because there is no HTT to worry about. No memory dividers to mess with in relation to the FSB (do you want high fsb and a divider or lower fsb and higher divider?)

 

[myocardia]

Originally posted by: Atheus
If you're not raising the voltages you won't damage anything. Even if you are raising voltage, it's pretty difficult to do actual permanent damage. Your motherboard is not designed to fry processors.

That's not true at all. No matter how good your cooling is, many motherboards are capable of supplying enough voltage to fry processors in minutes, if not instantaneously. My DFI Ultra-D, for instance, is capable of supplying 1.55v x123% maximum vcore, IIRC. That would be ~1.90v. Would you really like me to give your Socket 939 processor that much voltage?

Originally posted by: gersson
this reminds me -- we need a Quick and dirty guide to C2D OCing.

True, it would be nice to point all of these C2D OCing questions to a stickied, all-inclusive C2D OCing thread.

 

[cmdrdredd]

Originally posted by: myocardia

Originally posted by: Atheus
If you're not raising the voltages you won't damage anything. Even if you are raising voltage, it's pretty difficult to do actual permanent damage. Your motherboard is not designed to fry processors.

That's not true at all. No matter how good your cooling is, many motherboards are capable of supplying enough voltage to fry processors in minutes, if not instantaneously. My DFI Ultra-D, for instance, is capable of supplying 1.55v x123% maximum vcore, IIRC. That would be ~1.90v. Would you really like me to give your Socket 939 processor that much voltage?

Originally posted by: gersson
this reminds me -- we need a Quick and dirty guide to C2D OCing.

True, it would be nice to point all of these C2D OCing questions to a stickied, all-inclusive C2D OCing thread.

I still have no idea why some boards offer voltages that you can't even really use. Even if you live in antarctica and have a Phase cooling system 1.9v would probably kill it right away.

 

[F1shF4t]

Originally posted by: cmdrdredd

Originally posted by: myocardia

Originally posted by: Atheus
If you're not raising the voltages you won't damage anything. Even if you are raising voltage, it's pretty difficult to do actual permanent damage. Your motherboard is not designed to fry processors.

That's not true at all. No matter how good your cooling is, many motherboards are capable of supplying enough voltage to fry processors in minutes, if not instantaneously. My DFI Ultra-D, for instance, is capable of supplying 1.55v x123% maximum vcore, IIRC. That would be ~1.90v. Would you really like me to give your Socket 939 processor that much voltage?

Originally posted by: gersson
this reminds me -- we need a Quick and dirty guide to C2D OCing.

True, it would be nice to point all of these C2D OCing questions to a stickied, all-inclusive C2D OCing thread.

I still have no idea why some boards offer voltages that you can't even really use. Even if you live in antarctica and have a Phase cooling system 1.9v would probably kill it right away.

Coolling wont protect from that voltage. Also dfi ultra d can put 4 volts through ram, i wonder what ram will last at all on that voltage.

 

[cmdrdredd]

Exactly, I wonder if they put the option in because they wanted to see how many people called them with fried components so they could laugh at them.

 

[F1shF4t]

Originally posted by: cmdrdredd
Exactly, I wonder if they put the option in because they wanted to see how many people called them with fried components so they could laugh at them.

Braging right and marketing prolly, larger numbers look better lol.
But besides that the ultra d is an owesome overclocking board, easily one of the best for socket 939.

 

[cmdrdredd]

Originally posted by: Dark Cupcake

Originally posted by: cmdrdredd
Exactly, I wonder if they put the option in because they wanted to see how many people called them with fried components so they could laugh at them.

Braging right and marketing prolly, larger numbers look better lol.
But besides that the ultra d is an owesome overclocking board, easily one of the best for socket 939.

I own and use a LP UT SLI-D with an x2 3800+ running 2.6Ghz.

Unfortunately that super gaming mobo is running on the standard onboard sound and a 6200LE heh. I use that system mainly to encode a DVD or something while I'm gaming or can't stop what my C2D system is doing at the time. Sometimes I'll lookup something online without having to alt-tab out of a game or pause an encoding/burn.

It just seems wrong in a way to use this system in such a way. Although I'm going to start using D2OL on it pretty much 24/7 soon.

 

[Elfear]

You'd be surprised how many hardcore guys don't mind throwing crazy volts at stuff. The C2D has only been out a few months and guys are already giving it 1.7-1.8V. The 939 chips have been given 1.9-2.0V if I remember correctly. The high voltage options are there for the Xtreme guys.

 

[cmdrdredd]

Originally posted by: Elfear
You'd be surprised how many hardcore guys don't mind throwing crazy volts at stuff. The C2D has only been out a few months and guys are already giving it 1.7-1.8V. The 939 chips have been given 1.9-2.0V if I remember correctly. The high voltage options are there for the Xtreme guys.

There's nothing glorious about getting a CPU-Z Shot and a SuperPi 1M score before it fries. At least not to me.

 

[Conky]

Originally posted by: cmdrdredd

Originally posted by: Elfear
You'd be surprised how many hardcore guys don't mind throwing crazy volts at stuff. The C2D has only been out a few months and guys are already giving it 1.7-1.8V. The 939 chips have been given 1.9-2.0V if I remember correctly. The high voltage options are there for the Xtreme guys.

There's nothing glorious about getting a CPU-Z Shot and a SuperPi 1M score before it fries. At least not to me.

Exactly, if you burn it up you can't reproduce your benchmarks.

 

[cmdrdredd]

Originally posted by: Beachboy

Originally posted by: cmdrdredd

Originally posted by: Elfear
You'd be surprised how many hardcore guys don't mind throwing crazy volts at stuff. The C2D has only been out a few months and guys are already giving it 1.7-1.8V. The 939 chips have been given 1.9-2.0V if I remember correctly. The high voltage options are there for the Xtreme guys.

There's nothing glorious about getting a CPU-Z Shot and a SuperPi 1M score before it fries. At least not to me.

Exactly, if you burn it up you can't reproduce your benchmarks.

What is probably even more humerous to me is the SuperPi result wouldn't validate because the CPU gave an incorrect result

 

[F1shF4t]

Originally posted by: cmdrdredd

Originally posted by: Dark Cupcake

Originally posted by: cmdrdredd
Exactly, I wonder if they put the option in because they wanted to see how many people called them with fried components so they could laugh at them.

Braging right and marketing prolly, larger numbers look better lol.
But besides that the ultra d is an owesome overclocking board, easily one of the best for socket 939.

I own and use a LP UT SLI-D with an x2 3800+ running 2.6Ghz.

Unfortunately that super gaming mobo is running on the standard onboard sound and a 6200LE heh. I use that system mainly to encode a DVD or something while I'm gaming or can't stop what my C2D system is doing at the time. Sometimes I'll lookup something online without having to alt-tab out of a game or pause an encoding/burn.

It just seems wrong in a way to use this system in such a way. Although I'm going to start using D2OL on it pretty much 24/7 soon.

What can u do about it, i'm using my old northwood in my sig, as a box which just runs 24/7 mostly torrents and downloads, that chip served me well, but its getting old, it used to kick serious but back in the day.

Too bad i cant afford conroe yet, dont wanna hijack the thread... but how much is that conroe better than ur x2 (both overclocked), just general usage, playing games and encoding.

 

[gobucks]

as far as i know, overclocking does not substantially reduce the life of your chip unless you do some serious overvolting (> 15% or so) or let your temps get too high for long periods of time. As long as you can get your chip running prime95 stable at decent temps, I'm pretty sure you'll be fine. It's not like the chip is inherently designed to run at a given speed or voltage. Clockspeeds are arbitrarily set to meet market demand; most are quite capable of running at the same speeds as $1000 chips, but they lower the multiplier because most people don't want to spend $1K on a CPU. the voltage is simply what Intel or AMD decided was adequate to make the average chip run stable, but it isn't set in stone either (just look at AMD's san diego cores, some chips have different default vcore, depending on how good the die is, even though they are all built on the same process). There is nothing inherently damaging about running a chip faster than at stock speed, as long as you don't go overboard. Besides, if you're in the business of overclocking, you'll probably be upgrading every few years, so if your overclock knocks a year off of the lifespan of your chip, you'll probably be replacing it long before it dies.

I had an A64 3000+ overclocked to 2.55GHz (40% OC) for about a year and a half, and i didn't notice any stability problems with it whatsoever. I sold it about 6 months ago, and i haven't gotten any angry emails from the buyer as of yet, so i imagine he isn't having any problems with it either.

FYI, if you're looking at a Core 2 Duo, the generally agreed guidelines for safe overclocks of an e6300/e6400 are around 3GHz/1.4V if you have crappy cooling, and 3.2-3.3GHz/1.5V if you have good cooling (scythe infinity, tuniq tower, zalman 9500, etc.) Make sure to check for stability using prime95, and watch your temps with the CoreTemp utility - make sure you don't let your chip get above 60C - Intel says temps higher than 61.5C can damage your CPU long term. they are probably being conservative, but i wouldn't push it.

 

[Raider1284]

As far as I know, As long as you keep the cpu cold, it shouldn't be causing any damage..right? Heat is the killer. Increasing the voltage increases the heat, and therefore the damage; but if you keep it cool I don't see how it could cause damage. Please correct me if I'm wrong.

 

[F1shF4t]

Originally posted by: Raider1284
As far as I know, As long as you keep the cpu cold, it shouldn't be causing any damage..right? Heat is the killer. Increasing the voltage increases the heat, and therefore the damage; but if you keep it cool I don't see how it could cause damage. Please correct me if I'm wrong.

No the voltage will kill the cpu also, the transistors inside the chip will eventially get destryed from too much voltage, well from the current, but more voltage = more current since the resistance of the chip doe not change (well it does with temperature which causes more current to flow through the chip, so if u dont cool it the temp will just ramp up and destroy the chip, thermal runaway)

 

[Atheus]

Originally posted by: Dark Cupcake

Originally posted by: Atheus
If you're not raising the voltages you won't damage anything. Even if you are raising voltage, it's pretty difficult to do actual permanent damage. Your motherboard is not designed to fry processors.

If your not raising voltages, the extra clock speed will still damage the chip

Why? How? Do transistors have a certain number of 'switches' in their life and then they die?

 

[F1shF4t]

Originally posted by: Atheus

Originally posted by: Dark Cupcake

Originally posted by: Atheus
If you're not raising the voltages you won't damage anything. Even if you are raising voltage, it's pretty difficult to do actual permanent damage. Your motherboard is not designed to fry processors.

If your not raising voltages, the extra clock speed will still damage the chip

Why? How? Do transistors have a certain number of 'switches' in their life and then they die?

Cpus are made using gates based on cmos transistors...
Simply when they switch there is a small short circuit, thats when they use power. When they are static they draw almost no power. So as u increase the clock speed the switching happens closer and closer together resulting in more power consumption and heat.

So simply the faster the clock speed the more current u put through those transistors. And current slowly destroyes the little wires in the chips.
Anyways i just described just the very basics, ofcourse its a lot more complex when it come to todays cpus.

A good read