When OVERCLOCKING, how does an increased VOLTAGE harm the CPU?

x86

Banned
Oct 12, 2001
397
0
0
It is usually the case that in order to reach higher speeds the voltage has to be increased, but how does this damage the CPU?

Thanks

-DocSmarts

---

This is way off topic.

AnandTech Moderator


 

brainalien

Member
Feb 6, 2001
29
0
0
I believe that would be heat correct? The higher voltage will generate more heat in the processor, which can cause a-lot of problems, worst of all permanently damaging it.
 

Budman

Lifer
Oct 9, 1999
10,980
0
0
It might even catch fire & you'll have to put out the flames with a pitcher of water.:Q ;)
 

Lord Evermore

Diamond Member
Oct 10, 1999
9,558
0
76
Stick a 5V battery to your tongue. Stings a bit but doesn't damage you. Now stick your tongue in a 120V socket. You dead. Same thing happens when you increase voltage too much. Any circuit (with your body being part of the circuit when you touch it to electrodes) can only handle a certain amount of current before it burns out.

However with processors very slight voltage increases can be damaging because they're designed to run at such low voltages and currents. Generally a slight increase is not immediately damaging, as most processors are capable of working at somewhat higher voltages (and in fact that's how both AMD and Intel occasionally manage to tweak just a bit more speed out of an aging architecture; they rate the processor at something like 0.1 volts higher than the previous speeds, and that allows them to rate the clock speed a bit higher -- in the end, you basically get an already overclocked CPU from the manufacturer). However when you start increasing the voltage too much, it begins to damage the processor. It may not be immediately destroyed, but it can slowly degrade the CPU, and over time you might begin noticing more errors or instability. However, processors are designed to last 10 years or so at their "rated" speeds and voltages. Even if you cut the lifespan in half, that's 5 years, and anyone overclocking that much isn't likely to care if it lasts only even one year. It is of course possible to increase the voltage so much that you completely fry the CPU and sometimes take other components with it.

You can sometimes tell what's safe to do. Like with an Athlon t-bird, 1.85 volts isn't outside its operational range, but 1.75V is what they're rated at. AMD makes white-papers that show them operating at 1.85V, so it's usually safe to run them like that. Of course, at increased voltages you have increased heat, which compounds the issue and makes it more likely that you'll encounter errors or instability if you don't cool it well enough.
 

RaynorWolfcastle

Diamond Member
Feb 8, 2001
8,968
16
81


<< It might even catch fire & you'll have to put out the flames with a pitcher of water.:Q ;) >>



DocSmarts, I'm afraid you're doomed to have replies like these appear in all of your threads. If I were you, I'd change handles while you still have few posts ;).

This is NOT A JOKE, it is a serious recommendation if you want to avoid being harassed in the future.

-Ice
 

John

Moderator Emeritus<br>Elite Member
Oct 9, 1999
33,944
4
81
Is the funny MOD on duty?
rolleye.gif
:Q
 

RaynorWolfcastle

Diamond Member
Feb 8, 2001
8,968
16
81


<<

<< DocDumb@$$ >>



LMAO.
:D
>>



ROFL
HAHAHAHAHAHA...
That's just too much for me

At least 1 person took note of what I said

-Ice
 

ST4RCUTTER

Platinum Member
Feb 13, 2001
2,841
0
0
Lord E pretty much nailed it. As a circuit becomes smaller, and the paths closer together, the chance of electrons "finding the path of least resistance" becomes important. Increasing Vcore to a CPU doesn't have to increase core temperature to cause catastrophic failure. A simple breakdown of the silicon between the circuit paths due to excessive voltage can result in a CPU failure. This is especially relevant to today's CPU's due to the coming .13um process.
 

Ionizer86

Diamond Member
Jun 20, 2001
5,292
0
76
Whoa, when did DocSmarts handle change? Last time I was in the thread, his handle was still what it was, which is ironic for someone capable of burning up a CPU....
rolleye.gif
 

x86

Banned
Oct 12, 2001
397
0
0


<< Somehow I have a feeling the Doc won't be back. >>



WRONG!

This is just a little prank of an AMD Zealot that can't accept the fact that AMD chips are horrible. As for who did this: Get a life.

I am currently programming a MOD to rechange my user name.

-DocSmarts
 

x86

Banned
Oct 12, 2001
397
0
0


<<

<< Nope, a mod must've done it

-Ice
>>



LOL!
>>



I WILL GET THEM!

Thank you very much
 

x86

Banned
Oct 12, 2001
397
0
0


<< It's about time. He's been trolling too long. >>



SCREW YOU! For your information, I have not been trolling, just stating my opinion.
 

Hulk

Diamond Member
Oct 9, 1999
5,186
3,802
136


(In Yoda Voice) "Voltage leads to increased potential, increased potential leads to energetic electrons, energetic electrons leads to suffering for the semiconductor!"
 

AGodspeed

Diamond Member
Jul 26, 2001
3,353
0
0


<<

<<

<< Nope, a mod must've done it

-Ice
>>



LOL!
>>



I WILL GET THEM!

Thank you very much
>>



You can't get back at an AT moderator DocDumb@$$. He'll just destroy you. ;)
 

Ionizer86

Diamond Member
Jun 20, 2001
5,292
0
76
This is just a little prank of an AMD Zealot that can't accept the fact that AMD chips are horrible. As for who did this: Get a life.

Hmm... DocSmarts, your logic seems to imply that if you drive a Brand X vehicle, and you crash it, it is the fault of Company X, who made a vehicle in such a manner that it is crashable due to user error.

In all honesty, if you purchase another 8KHA+ and Athlon XP 1700+, it will total around $200, and you will still be ahead compared to buying a Northwood 2.Aghz and a 8IRXP in the first place. They're inexpensive compared to intel setups, so even if you burn one chip and board, you simply get another.
 

x86

Banned
Oct 12, 2001
397
0
0


<<
You can't get back at an AT moderator DocDumb@$$. He'll just destroy you. ;)
>>



Oh really? Why is that?