The Overclocking Limiter

[Gladius]

What is it exactly that makes it so that you cannot just go faster and faster? what makes the system unstable? Is it lack of power? in which case you need a different PSU. Or is it heat? in which case you need better cooling. Or what? Or is it all your hardware as a whole, where a certain combination can handle all the power better? beats the heck out of me.

I have a 1.4 GHz Athlon and I can only get it stable at 1.466... im hoping heat is my problem, cause I can always get a new case and more cooling.

------------------
***
My System Specs

[mrspm2003]

heat is the main problem. Heat is lost energy, so the energy going at your processor at fast speeds tends to leak at a faster rate if its going to fast, therefore creating heat. There are many elements that prevent you from o/c higher, power, heat, fsb, all limit your speeds. Sorry if i dont make any sense, im sleepy right now

------------------
My 2 rigs

[retardedtimmah]

i am not aware of the EXACT limting factor but i do know that it has to do with cpu signal interference just not being able to generate the frequency that you set via jumpers or the bios...and it also starts making errors/miscalculations...i am not sure if this is due to excessive heat output...i know i've read before about why cold cpu's can go faster...i forget though...dang i'm sorry...? anyone know?
-R. tImMAH

------------------
Re·tärded Tim·my
----------------
re·tard1 (r-tärd)
v. re·tard·ed, re·tard·ing, re·tards
v. tr.
To cause to move or proceed slowly; delay or impede.

[sapasion]

Originally posted by Gladius:
What is it exactly that makes it so that you cannot just go faster and faster? what makes the system unstable? Is it lack of power? in which case you need a different PSU. Or is it heat? in which case you need better cooling. Or what? Or is it all your hardware as a whole, where a certain combination can handle all the power better? beats the heck out of me.

I have a 1.4 GHz Athlon and I can only get it stable at 1.466... im hoping heat is my problem, cause I can always get a new case and more cooling.

PSU, cpu voltage and cooling are the first three areas I'd look at.


------------------
OC Crusaders
My Humble Rig

[Gerbil]

There are little electrons moving around inside the CPU. It's kind of like being a race car driver. When you overclock, you are taking more traction away from their tires, but they keep going faster (this is what the heat does). Some of them wreck, and it's no big deal. When many wreck, the system crashes, locks up, or whatever. More cooling allows the electrons to drive better.

You can't overclock infinitly though, because the heat is still generated. You can still remove the heat and therefore keep the CPU cool, but that doesn't mean that the heat won't have already affected some of the electrons.

That's how someone explained it on another message board a few days ago.

He said more voltage and stuff gave them more "traction".

------------------
---> My System's <---

Abit KT7A
Athlon XP 1600+
GeForce 2 GTS 32mb
Crucial 256mb PC133 CL2
SB Live! Value

[This message has been edited by Gerbil (edited January 07, 2002).]

[Gladius]

cool thanks a bunch everyone.

Well I think I definitely need more cooling, because before I even tried to OC my cpu temp was around 56 degree C. I know that is hot and I know why it is so hot.. bad HSF and a mid tower case with neon lighting :P

------------------
***
My System Specs

[Dan792]

Originally posted by mrspm2003:
heat is the main problem. Heat is lost energy, so the energy going at your processor at fast speeds tends to leak at a faster rate if its going to fast, therefore creating heat. There are many elements that prevent you from o/c higher, power, heat, fsb, all limit your speeds. Sorry if i dont make any sense, im sleepy right now

Actually, heat is a transfer of energy. According to the Law of Conservation of Energy, energy cannot be created or destroyed. Sorry, couldn't resist. :P



------------------
"Let every nation know, whether it wishes us well or ill, that we shall pay any price, bear any burden, meet any hardship, support any friend, oppose any foe, in order to assure the survival and the success of liberty."
~ John F. Kennedy, Inaugural Address, 1961

[superjohnny]

A lot of times it can be the motherboard, memory, or a PCI device. Once you have the right power-supply and cooling there are a couple of things that can hold you up though. When you increase the FSB you increase the memory, PCI, and AGP busses so I have known people who couldn't overclock because their video card was holding them up. I know many times SCSI controllers don't like a over-spec PCI bus. Cheap RAM can often be the culprit.

Your best bet to get a good overclock is to ask around about the best parts and learn from the failures of other people. Some chipsets and motherboards don't like overclocking as well as others. I remember the KT133A chipset was supposed to be awesome and the Abit KT7A was supposed to be the best overclocker around... mine wouldn't get past 142MHz fsb I was pissed. I tried replacing the memory, video card, processor, I took out all the PCI cards... nothing worked. I finally just accepted that that motherboard didn't clock as well as others (like the Iwill KK266).

Ask around and don't be afraid to spend another 10-15% on parts to get the very best.

------------------
My stats @ OCgurus

[Option_Man]

Sometimes you get a chip that wont overclock. It has to do with the quality of that particular chip. Most chips can go faster than they are rated, but sometimes the are rated just right. I had two celerons that I bought at the same time. I clocked one at 200 mhz over and the other wouldnt over clock. I am not saying that that happened to you, but it is possible. The first overclock I did took me 2 weeks to get right. I had to read many forums and try different combinations of speeds and voltages. I also remember kicking a big dent in my file cabinet at about 4 in the morning.

[wenus2]

Originally posted by Dan792:
Actually, heat is a transfer of energy. According to the Law of Conservation of Energy, energy cannot be created or destroyed. Sorry, couldn't resist. :P

ahhh, but it can change form...

------------------
Athlon XP 1900+
Alpha Pal8045
Abit KR7A RAID
2x256Mb PC2100 /w ECC
Maxtor plus 60gb ATA-133
Seagate Barracuda 40Gb ATA-100
Cardex GeForce3
SB Live X-Gamer
Plextor 24x10x40
Toshiba DVD 16X/48X
Enermax case /w 430W EG465P-VE

[talldude]

Originally posted by mrspm2003:
heat is the main problem. Heat is lost energy, so the energy going at your processor at fast speeds tends to leak at a faster rate if its going to fast, therefore creating heat. There are many elements that prevent you from o/c higher, power, heat, fsb, all limit your speeds. Sorry if i dont make any sense, im sleepy right now

In this particular context, he's clearly stating that energy is lost from the system...and transformed to heat which is dispersed out from the core.

This is how i see your overclock is limited, in order:
lack of power
excessive heat
current leakage

The last one is really the one that limits extreme overclocking (providing you have superb cooling). It causes interference within the core which leads to miscalculations. This leads to crashes of course. It can be countered to a degree by increasing the voltage to the core as the signals increase in strength. However, at one point or another, too many electrons simply leak out of their intended pathways which locks up the system.

The reason cooling helps keep the electrons on their intended circuits is that as things get colder, they get more conductive (lose resistance to allow for easier moving of electrons.) I believe this is contributed to the fact that metals have a sort of "sea of electrons" in the structure. They're not in "orbits" like in normal atomic structures, but rather floating around free between the metal atoms. When things are cooled down, they slow down. As a result, the electrons in the metal structure move less so they're more organized, and can be moved more easily from point A to point B as there are less random collisions. (that's just the way I see it from physics and chem background).

Lack of power is obviously going to stop OC attemps...so no need to explain that.

I hope that's informative to you guys, correct any mistakes you see.


------------------
squire of an OC Crusader
Abit SH6 (finally working )
p3 slot1 700@860mhz (cooled by alpha p3125)
Max OC to boot in windows 896 mhz
creative geforce 256 120@150 core 166@200 memory
soundblaster live platinum
2x 128 mb pc-100 ram cas2 (awesome stuff)
28.5GB barracuda ATA
creative CDRW 4x4x32
buffalo 11mbs wireless lan
SE Overclockers SETI@Home Team

[Dan792]

Originally posted by wenus2:
ahhh, but it can change form...

Very true...


------------------
"Let every nation know, whether it wishes us well or ill, that we shall pay any price, bear any burden, meet any hardship, support any friend, oppose any foe, in order to assure the survival and the success of liberty."
~ John F. Kennedy, Inaugural Address, 1961

[KJY9]

Here's my take on it:
#1) Voltage (needed to OC)
#2) Heat (as a result from higher voltages)
#3) PSU (not being able to keep up the rails at voltages needed for OC)
#4) Extreme Voltages (above 2.0 VCore)
#5) Extreme Cooling (needed for extreme Voltage)

Just my .02 from experience

------------------
Take this link to send mail to our boys in Afghanistan.
http://www.anyservicemember.org/
KJY9's Rigs
Whoever said "ignorance is bliss"??
Compare 3DMark