Is underclocking a dangerous thing?
Forum Overclocking : General Discussions - Is underclocking a dangerous thing?
i have an big arguement with a person and he said by underclocking the CPU it will post harm to the motherboard.because what i said is running a CPU at 100x6 with lower vcore,but everything on normal.but he said the NB will suffer as the NB is running a higher(normal) voltage thats actually required for the speed.i said back to him the voltage on the NB is still running with in spec even though the bus speed isnt.i told him that only overclocking and overvolting will cause possible harm to the parts.by running them lower is fine.
so what you guys think.that running at lower speed(UNDERCLOCKING) is dangerous like overclocking?and will cause damage as such?
basically his point is running excess(stock) voltage for lower speed(out of spec) is a dangerous.
and my point is the chipset itself is running on stock voltage and its fine even if the bus speed 400mhz is running lower then the state natively supported 533-1333mhz
Message edited by iluvgillgill on 05-03-2008 at 08:23:24 PM
I'd say your right. Your motherboard runs at a designed voltage, it doesn't matter how low the bus speed goes it still requires the same voltage to create the frequency.
On a CPU you under clock and lower the voltage to save power and limit the heat. I'm guessing you could under volt the northbridge slightly, but i think it'd need the stock voltage to carry out it's normal job of talking to the FSB, RAM, PCI-E and then to stay in contact with the southbridge.
However, I'm no expert maybe someone else can talk more technical jargon and make sense?
thank you close_deal.thats exactly my point.i also pointed out that if a chipset is rated to run at 1333 natively then it doesnt mean anything lower is impossible or dangerous as he said.
im sure you cant undervolt any parts apart from CPU and the rest on "Normal" or "default"
Who is this "guy" you're refering to?
Reply to dagger
well he would post in here sooni think.beginning with "E".im off to work now
hehe that aint to hard to work out...
| iluvgillgill wrote : well he would post in here sooni think.beginning with "E".im off to work now |
Poor guy, working on a saturday.
Who's E?
How did he come up with that anyway? People are weird.
Message edited by dagger on 05-03-2008 at 08:59:22 PM
Reply to dagger
Ends with 'giri' i'm guessing
. Tiz a weird thing to argue about , but still it beats clock watching for my new case
.
Ah are you referring to Evilonigiri and his underclocked Q6600? ![]()
| Evilonigiri wrote : I think I lost already...I have my Q6600 running at 2.0GHz... We should have an underclocking contest. |
| njalterio wrote : @ Evilonigiri.....why is your q6600 underclocked?? |
| Evilonigiri wrote : Didn't need the extra power, and also I have a crappy cooler. I learned something though. Vcore for 2GHz is .9875v, so much lower than what is needed for 3.15Ghz. |
http://www.tomshardware.com/forum/ [...] w=0&nojs=0
Message edited by Shadow703793 on 05-03-2008 at 09:25:13 PM

Reply to Shadow703793
Hehe thats where I guessed it from
.
I don't see why it should. I guess there is only one way to find out, though
Antec 300 Illusion > Asus 4850 512MB w/AM Cooler > Corsair 650HX > CM Hyper 212 Plus > Other: PII 940/Gigabyte 790GX/4850 1G
Reply to EXT64
Underclocking a 1066 FSB processor is theoretically the same as running an 800FSB processor. No harm in that for sure.
Message edited by boonality on 05-03-2008 at 11:14:24 PM
Oh ya, email is the heart of work.
Reply to boonality
| Shadow703793 wrote : Ah are you referring to Evilonigiri and his underclocked Q6600? |
lol nice guess.
| amdfangirl wrote : Underclocking is by no means dangerous. Undervolting is by no means dangerous. Its just unstable.
|
thats what i pointed out to him.by running a lower speed does not mean dangerous by any mean.and lets put the performance degrade on the side.but he use 3ghz@1.7V being dangerous.yes to that case what he said is correct,but when underclock the NB/FSB to lower but with stock voltage wouldnt cause a harm because that chipset is design to take that voltage regardless what speed as long as its stable(of course more voltage for higher speed).
i wonder why he still not post in this thread yet.probably we all proved him wrong.lol
thank you again!
i really want that friend to see what all you guys said.that guy is funny he can side track himself and talk as in manufacture's view and said should not run parts out of spec even lower then and so that would be dangerous.
i feel like slapping him,because i told him we are consumers why are you talking in manfacture's view?as we are not going to benefit from it or something.and he just like just put me as a manufacture!
I MEAN WHAT'S THE POINT OF THAT?think like a manufacture?
Shadow703793
and he is the actual person denying what he is doing.underclocking his Q6600.
Then again, why not just spend the money and get a better cooler?
all CPU and ram is the same in the same model.but the manufacture just select a lower multiplier for them so they run slower and sell for cheaper as well.
so a q6600 is a downclocked qx6800 but with locked lower multiplier.but the CPU it self is CAPABLE OF RUNNING AT 2.93GHZ because its the same chip as the more expensive brother.and once again he point it that as dangerous as its running outside manfactures stated spec.i guess he never think its actually running in spec for the core inside the wrapping.
^ read again!!!><
basically he said run CPU/Mobo/Ram faster or slower(out of spec)is dangerous even though the actual partinside it is the same.but with multiplier difference or different stock speed.
get it now?
yeah.normal the lower speed parts have lower level in quality as well.manufacturing variation and defect.couldnt run as high as the others.but the chip when its made.its design to run at the same voltage.but the actual silicon inside disallow it to run at a higher speed even with the same "normal" voltage.that why when we OC a cpu we put more voltage to offset or overcome the defects or inferiority of the CPU.
is INFERIORITY a word?i think i just made it up.lool
I agree, under clocking won't damage anything.
You should never run anything out of the manufacturers spec. Listen closely to what you are commanded to do and comply.
Resistance is futile you must comply. ![]()
Except the Q6600/P35 they run at an easy 3G.
Message edited by Zorg on 05-04-2008 at 07:55:08 AM
so reply to his question.underclocking to 100(QDR400) where a board support only 133(QDR533) Pentiums.so that can be look as out of spec.as the NB is using the stock voltage rated for 533mhz and over.
you think its dangerous?
If the mobo supports the chip and has a setting for 100 then no.
yes to both.
just to let you guys know how this arguement started is wheni just talking some funny stuff to him,i sed if i run the e2160 at 100x6 i wonder what the setting and tmep would be.and he just started talking about its being dangerous!lol
yeah i know.its possible that why i noted that to him and i said i will try it out.but he go and try to scare me away!lol
Scare me away? *Scoffs*
Of course I won't run away. I was busy.
So here goes my part of the argument:
What I tried to say was 100Mhz FSB is dangerous simply because the voltage for it is too high. For example, if 333MHz requires 1.3V, the lowest voltage possible for the motherboard, and 100Mhz requires only 1.0V, which isn't possible on the motherboard, setting to 100Mhz will overvolt the motherboard. As we all know, overvolting is not good.
As for dangerous, I did stress to iluvgillgill that since 100Mhz is out of the manufacturer's specs, it's "dangerous", just like overclocking to 400Mhz FSB is. Apparently he missed that part, conveniently.
*Lesbian Lover Club* - founder Assman
Reply to Evilonigiri
I thing it goes like this:
Underclocking - is fine for power saving, less heat and prolonging lifespan, but it may cause instability if you go too low. Not dangerous. Comp goes sloow.
Stock - well we all know that works right
. Boooring who does that...
Overclocking - dangerous because your making it work out of it's boundaries both voltage and Mhz wise... Increasing heat, power consumption and shortening lifespan. More raw Power and bigger e-penis.
| closed_deal wrote : I thing it goes like this: Underclocking - is fine for power saving, less heat and prolonging lifespan, but it may cause instability if you go too low. Not dangerous. Comp goes sloow. Stock - well we all know that works right Overclocking - dangerous because your making it work out of it's boundaries both voltage and Mhz wise... Increasing heat, power consumption and shortening lifespan. More raw Power and bigger e-penis. |
The e-penis is important. Make it grow, then poke ppl with it...
Reply to dagger
| closed_deal wrote : I thing it goes like this:
|
Excellent point, but my argument is, is underclocking safe when overvolted? I don't think that will prolong lifespan.
I created a new thread here: http://www.tomshardware.com/forum/ [...] unsafe#bas
I guess I'm an one man army. What OCing enthusiast will UC?
*Lesbian Lover Club* - founder Assman
Reply to Evilonigiri
| Evilonigiri wrote : Excellent point, but my argument is, is underclocking safe when overvolted? I don't think that will prolong lifespan. I created a new thread here: http://www.tomshardware.com/forum/ [...] unsafe#bas I guess I'm an one man army. What OCing enthusiast will UC? |
It's not overvolting. The chip can take what it's rated to take, regardless of the frequency.
Reply to dagger
| dagger wrote : It's not overvolting. The chip can take what it's rated to take, regardless of the frequency. |
That's true, and I know that, but I don't think 800MHz at 1.5V will do any good[/extreme example].
*Lesbian Lover Club* - founder Assman
Reply to Evilonigiri
^^^Agree with dagger. The chip is ment to run with whatever voltage it takes. Put it in simple understanding as a 'bulb'.
The more you pump into it the brighter it gets, but too much can 'blow' it up. If you feed it too little it's just a dim light. Or a radio if you wan to include the frequency.
Message edited by closed_deal on 05-04-2008 at 08:45:17 PM
Does anyone know where the clock generator is for the NB? As in is it actually built into it. It is valid for the discussion.
Message edited by closed_deal on 05-04-2008 at 08:47:54 PM
| closed_deal wrote : ^^^Agree with dagger. The chip is ment to run with whatever voltage it takes. Put it in simple understanding as a 'bulb'.
|
I understand that perspective. Now using your example, would a lightbulb at 50% last longer than a lightbulb at 100%? Assuming that the voltage for each is the same, which probably isn't.
*Lesbian Lover Club* - founder Assman
Reply to Evilonigiri
| Evilonigiri wrote : I understand that perspective. Now using your example, would a lightbulb at 50% last longer than a lightbulb at 100%? Assuming that the voltage for each is the same, which probably isn't. |
It'll be obsolete long before it burns out anyway.
Reply to dagger
| dagger wrote : It'll be obsolete long before it burns out anyway. |
Yep, and we get those mercury-filled, power saving light bulbs. Oh and when you dim that light bulb, it blows up. I wonder why...
*Lesbian Lover Club* - founder Assman
Reply to Evilonigiri
It would because electrical devices only draw the Current they need. I can see your point here, because too much voltage can kill. What your suggesting though, the stock voltage wouldn't be enough to actually damage the chip.
| closed_deal wrote : It would because electrical devices only draw the Current they need. I can see your point here, because too much voltage can kill. What your suggesting though, the stock voltage wouldn't be enough to actually damage the chip. |
Well, using the chip is already damaging.
I'm just wondering if it'll be more damaging than doing stock. If it is, it proves my point.
*Lesbian Lover Club* - founder Assman
Reply to Evilonigiri
Ok i think i've got this right in my head now... A CPU set at 500Mhz can be forceably overclock by manually setting the frequency you want it to run at an give it a little more voltage. For the clockgen on the NB to keep up you have to increase the voltage slightly. Thats why you get FSB wall on the mobo first normally, becuase the CPU has a little more headroom than the NB can handle.
Therefore decreasing the actual frequency of the of the CPU will lower whats travelling through to the NB, but the NB itself is still within it's parameters because the FSB is now lower than it's standard running frequencies on stock voltage. However if the frequency gets to low for the NB clock gen (normally not the CPU) to handle you get the instability.
Ok i think thats right maybe a few more edit's to get this completely right
...
Message edited by closed_deal on 05-04-2008 at 09:16:55 PM
| closed_deal wrote : Ok i think i've got this right in my head now... A CPU set at 500Mhz can be forceably overclock by manually setting the frequency you want it to run at an give it a little more voltage. For the clockgen on the NB to keep up you have to increase the voltage slightly. Thats why you get FSB wall on the mobo first normally, becuase the CPU has a little more headroom than the NB can handle.
|
Interesting. You sure know a lot!
So is this more damaging, even though it may be a little, than at normal conditions?
*Lesbian Lover Club* - founder Assman
Reply to Evilonigiri
I would hazard a guess and say so. The NB chip is still running at it's req'd default voltage but drawing less current.
*EDIT* Also as 4745454b says the transistors (nand gate switches) even on micro processor level need a req'd voltage to open and close.
Message edited by closed_deal on 05-04-2008 at 09:43:18 PM
| closed_deal wrote : I would hazard a guess and say so. The NB chip is still running at it's req'd default voltage but drawing less current.
|
Although I would like to say I'm glad someone sees my point of view, it's not nearly enough to prove my point or iluvgillgill's.
Is there anyone else with some more proof?
*Lesbian Lover Club* - founder Assman
Reply to Evilonigiri
My electrical engineering only goes as far as aircraft avionics, so i can't give solid facts just slightly educated ones
. Just the way the CPU and NB talk seems to be similar to the way the Main comp sends data packets to the rest of the systems...
Message edited by closed_deal on 05-05-2008 at 12:36:43 AM
| closed_deal wrote : My electrical engineering only goes as far as aircraft avionics, so i can give solid facts just slightly educated ones |
Well, you know more than me.
Wow, EE. I'm impressed.
*Lesbian Lover Club* - founder Assman
Reply to Evilonigiri
Nah it's not that impressive. My qualification doesn't even allow me to do basic housing rewiring tiz daft , but hey I can work on aircraft so i can't be 'that' dangerous hehe.
I'm stuck between Pre-Engineering Physics and Electrical Engineering for my major. There's only 26people for Pre-Engineering Physics and nearly 600 for EE, so I'm wondering if I should switch to EE
Then I can answer my own question and end this discussion once and for all.
*Lesbian Lover Club* - founder Assman
Reply to Evilonigiri
Lol - EE is a easy really. Took 14 months for me to be fully trained from a 16year old squirt to a 17 yo slightly taller squirt
. Plus after 6 years on the job it tends to get blown out and replaced by the cobwebs hehe.
P.S If you ever do wave guide theory and actually retain the information give me a call b'coz that almost turned my world unside down
Message edited by closed_deal on 05-05-2008 at 12:50:05 AM
Nice, if I ever do, I will give you a call.
*Lesbian Lover Club* - founder Assman
Reply to Evilonigiri
Evilonigiri, the mobo won't be damaged by lowering the frequency. As was said, you aren't over-volting the CPU, RAM or mobo, just lowering the speed.
Let it go.
Message edited by Zorg on 05-05-2008 at 04:47:15 AM
There are 11 identified and unidentified users. To see the list of identified users, Click here.
Please mind
You are about to answer a thread that has been inactive for more than 6 months.
If you still wish to proceed, please ensure that your posting is original and does not duplicate or overlap any prior responses to this thread.
