Extreme Overclocking???

[Superman53142]

Theoretically, if you could pump enough juice into the cpu, and had some monster cooling solution, couldn't you increase both the FSB and multiplier to their highest overclockable settings. For instance, people get their T-birds to 180mhz FSB. And others get a 12 or 12.5 multiplier on their 1.4Ghz t-birds. If you could run a t-bird at 180 FSB with a 12.5 multiplier you'd have 2250mhz. What factors are limiting this kind of extreme overclocking?

[Ferg]

This has been done with liquid nitrogen. The cpus last long enough to post in windows then die along with usually half of the components in the computer.

[Hoot]

Above about 150 Mhz FSB, you have to consider the other devices on the PCI/AGP/ISA busses, not to mention the RAM. Each one of those devices has a limiting speed, besides the CPU. It's quite a juggling act.

Hoot

[asmodean]

Components don't necessarily die when using liquid nitrogen. They have a high tendency to do so, but it doesn't always happen.

[Superman53142]

Hoot, they've got that new BIOS for the 8k7a that allows the 1/5 pci divisor. that elimates problems with the pci and agp. Memory is your only limiting factor. PC166 will do 180mhz.

[Fjeld]

Ferg (Jul 12, 2001 08:58 a.m.):
This has been done with liquid nitrogen. The cpus last long enough to post in windows then die along with usually half of the components in the computer.

Haha!! You said the right thing!!! There´s always a limit..... :)

[castle lager fan]

What holding back extreme overclocking?!
Dare I call it physics!!
First understand the makeup of a transistor. It takes time for the electrons to create a sufficient flow between the base and emitor to actively switch on the transistor. And then it takes time for that to get a sufficient flow through collector and emitor to actually do something usefull with it.

Now consider the following the more voltage potential you put on it the more force there is to push the little electrons around. SO they should go faster! WRONG they will still travel at the same speed. That one of the reason why the silicon technology used in CPU today is so tiny. Make the distance less for travel, so things will happen quicker! I doubt whether we will ever see 2 GHz CPU thats relaible with 0.18 mcron technology. WHy do you think the tech companies spend billions on developing making smaller transistors.

The reason why you want to raise the voltage on a transistor is to make it more stable at switch on. It will not make it switch on "faster" (some might disagree). It does make it more stable when you switch in on and off a lot. Like in clocking our little CPU's.

SO now you might ask what has it got to do with overclocking. Well CPU consist of millions of transistors as the basic building element and you can't go faster than the basic element. Oh I forgot heat creates leaks in silicon such as diodes and tramsistors which makes its isolation or switching capabilities less reliable. And the heat within a transistor is mostly created in the time it takes to switch on or switch off. It produces far less heat once it is fully on or fully off. That why you have the conflict of onterest of the faster I switch it (Clocking on CPU) the more heat it will produce and the less stable it will become. SO I up the voltage to make it more stable, but it produces more heat till he point where the silicon junction fails (CPU smoking!) SO I have to get rid of the heat (CPU cooling). Now the faster I switch the more voltage required to stabilise, the more heat I need to absorb from transistor.

Thus if I make CPU cooler I can clock faster without increasing voltage, because I have less leaking. More voltage I can go even faster, but more heat needs to go "away". SO bottom line is that the silicon structure in the current chips do have a limit. No matter how much voltage or cooling you apply you will not go beyond a certain limit!! To go faster the charactirstics of the silicon transistors have to be changed to work. That why we will get 0.13 micron (millionth of meter) chips soon oops already have some. Then 0.10, etc. some additional tricks on the compound SOI silicon on isolator.

OH, I forgot now we have to get into EMI properties within silicon chips and mobos ...........:-);-)

Hope it helped

[Warlord2]

you said a mouth full