Curse

Techie · Nov 08, 2005, 04:47 AM // 04:47

Well my friend had an old PIII machine sitting around with the following:

Pentium III - 500 Mhz
RAM - 64MB PC133
ATI Rage 8MB Video Card

So yeah I decided to bring over my dry ice and cathodes and containers, and we did an OC party! We managed to get a cheap case from a local store and swap the parts over. I added some case fans, about 3 x 120mm, and then I setup my dry ice. The idle temps stock were -58C, and load were -52C. We managed to pull off 2.8, yes, 2800 Mhz on a PIII! Anything past that didn't even POST, plus we didn't have high enough voltage

But I managed to get a 2300 Mhz OC off of a PIII with dry ice.

Oh and at 2.8 Ghz the idle temps were 3C and load temps were 14C. I love it!

Dry ice pwns!

Unfortunately, we couldn't do any benchmarks because the video card wouldn't run 3DMark01/03/05 and we didn't have any extra PC133 RAM sitting around.

EchoSex · Nov 08, 2005, 04:49 AM // 04:49

Quote:

Originally Posted by Techie

Well my friend had an old PIII machine sitting around with the following:

Pentium III - 500 Mhz
RAM - 64MB PC133
ATI Rage 8MB Video Card

So yeah I decided to bring over my dry ice and cathodes and containers, and we did an OC party! We managed to get a cheap case from a local store and swap the parts over. I added some case fans, about 3 x 120mm, and then I setup my dry ice. The idle temps stock were -58C, and load were -52C. We managed to pull off 2.8, yes, 2800 Mhz on a PIII! Anything past that didn't even POST, plus we didn't have high enough voltage

But I managed to get a 2300 Mhz OC off of a PIII with dry ice.

Oh and at 2.8 Ghz the idle temps were 3C and load temps were 14C. I love it!

Dry ice pwns!

Unfortunately, we couldn't do any benchmarks because the video card wouldn't run 3DMark01/03/05 and we didn't have any extra PC133 RAM sitting around.

A PIII.... at 2.8..... What the HELL is wrong with you boy?! You're gonna set the house on fire o.o

Otherwise, bravo. I thought I did good with my Radeon 9800XT, but you've taken the cake. I should look into dry ice rather than liquid cooling.

nightrunner · Nov 08, 2005, 05:49 AM // 05:49

O_O come set up a cooling system for my machine!!!

Techie · Nov 08, 2005, 04:23 PM // 16:23

Yeah dry ice is very inexpensive, but making the covers and containers is the BIGGEST pain in the #%!. Liquid cooling is very good for medium-high overclocks. We did eventually see smoke from the CPU so we turned the frequency down. It has a small burn mark but amazingly still works ok

Oh and btw what did you get your 9800XT to?

zehly · Nov 08, 2005, 05:05 PM // 17:05

That's a very nice idea there, unfortunately it doesn't work the way you said, which means you fabricated the story.

Dry Ice:

is really just carbon-dioxide in a solid state.
is always being sublimated, meaning that it cannot be stored in a sealed container.
must be very well ventilated, because carbon dioxide gas interferes with the ability of the body to take in the proper amounts of nitrogen and oxygen.
is rarely used in conjunction with electrical components. Liquid nitrogen works better.

The presence of carbon dioxide gas is normally around .05% of the atmosphere that we breathe in; we breathe out around 4-5%. Inhaling more than 5% of carbon dioxide gas is hazardous. Excess carbon-dioxide changes the pH of our blood and causes hemoglobin to function adversely.

Silicon-based microprocessors:

T-Junctions as well as other parts inside transistors react to change in temperature.
Change in temperature will affect the resistance across a circuit.
If a processor gets too cold, it conducts more electricity (resistance falls), and will eventually lead to a short/spike in electrical activity.
If a processor gets too hot, it conducts less electricity (resistance rises), and will eventually melt the dye, causing a short, and then fire.
Room temperature is relatively ideal for semiconductors in home-computing.
The Pentium III processor CANNOT EVER reach a temperature under -40C without destroying it. This is either powered on OR off.
You would not have a deficiency in voltage; you would have had too much!
The P3 CPU does not support 2.8GHz of oscillation. Period

So, all in all, what you did cannot be done. Oh, and this was proven too; I had my students try it in the lab using PROPER EQUIPMENT! The die shorted, sparked, and fizzled.

I'm not posting this as a flame, though, just as a minor correction.

~zehly

kg_lildude1 · Nov 08, 2005, 05:20 PM // 17:20

go zehly FTW! i declare techie's post officially pwned
=======================================
um i think my story qualifies not as a overclock but a meltdown i guess?

one fourth of july night while my whole family was wasted from my uncles homemade brew, we started it off with some bottlerockets in our bottles, then we um... started improvising, so then pots and vases started to um.... spontaneously combust?

after we used all of our fireworks, we still wanted more,.... hmmm what to do?

well we had a useless computer and we were gonna get a new one anyway so we decided to put it into overdrive

yea uh,... needless to say we got our fireworks >_>

lesson learned? never get wasted and want fireworks, then decide to get creative

Xue Yi Liang · Nov 08, 2005, 06:15 PM // 18:15

Quote:

Originally Posted by zehly

That's a very nice idea there, unfortunately it doesn't work the way you said, which means you fabricated the story.

Dry Ice:

is really just carbon-dioxide in a solid state.
is always being sublimated, meaning that it cannot be stored in a sealed container.
must be very well ventilated, because carbon dioxide gas interferes with the ability of the body to take in the proper amounts of nitrogen and oxygen.
is rarely used in conjunction with electrical components. Liquid nitrogen works better.

The presence of carbon dioxide gas is normally around .05% of the atmosphere that we breathe in; we breathe out around 4-5%. Inhaling more than 5% of carbon dioxide gas is hazardous. Excess carbon-dioxide changes the pH of our blood and causes hemoglobin to function adversely.

Silicon-based microprocessors:

T-Junctions as well as other parts inside transistors react to change in temperature.
Change in temperature will affect the resistance across a circuit.
If a processor gets too cold, it conducts more electricity (resistance falls), and will eventually lead to a short/spike in electrical activity.
If a processor gets too hot, it conducts less electricity (resistance rises), and will eventually melt the dye, causing a short, and then fire.
Room temperature is relatively ideal for semiconductors in home-computing.
The Pentium III processor CANNOT EVER reach a temperature under -40C without destroying it. This is either powered on OR off.
You would not have a deficiency in voltage; you would have had too much!
The P3 CPU does not support 2.8GHz of oscillation. Period

So, all in all, what you did cannot be done. Oh, and this was proven too; I had my students try it in the lab using PROPER EQUIPMENT! The die shorted, sparked, and fizzled.

I'm not posting this as a flame, though, just as a minor correction.

~zehly

You mean my liquid nitrogen cooled TRS-80 setup isn't going to work either?
Dang. Back to the drawing board.

Well put.

EchoSex · Nov 08, 2005, 08:16 PM // 20:16

Quote:

Originally Posted by Techie

Yeah dry ice is very inexpensive, but making the covers and containers is the BIGGEST pain in the #%!. Liquid cooling is very good for medium-high overclocks. We did eventually see smoke from the CPU so we turned the frequency down. It has a small burn mark but amazingly still works ok

Oh and btw what did you get your 9800XT to?

480 at the moment - the liquid cooling is a big part of that. I had to remove the heat sink and apply the surface directly. That copper plate just gets way too hot. I DID have a problem with "jaggies" before with it, not jaggies in the sense of jumbled pixals, but in the sense of the screen litterely breaking up. Ever since the newer Catalyst drivers have come out, no problems. I'm thinking of kicking it's ass to 500 when I have the two nVidia's I'm replacing it with handy. Just to see if it's possible.

Btw, it IS possible to kick a PIII's ass that high. It's called third party software, people. You can make a chip run on just about any speed you want - the trick is keeping it that way without it crapping it out. Don't belive me? Go find one of those old ass IBM processors and experiment. Hell, I've seen some pretty cool stuff done with computers in a minature freezer.

Techie · Nov 08, 2005, 09:13 PM // 21:13

Well would you like to see screenshots? I am telling you that dry ice keeps temps in the negative celcius, and the highest it ever got before it stopped posting was about 30F.

zehly · Nov 08, 2005, 09:56 PM // 21:56

Quote:

Originally Posted by EchoSex

Btw, it IS possible to kick a PIII's ass that high. It's called third party software, people. You can make a chip run on just about any speed you want - the trick is keeping it that way without it crapping it out. Don't belive me? Go find one of those old ass IBM processors and experiment. Hell, I've seen some pretty cool stuff done with computers in a minature freezer.

Third party software, unfortunately, cannot change the thermodynamic properties of any open or closed system. Sorry, not happening. Don't get me wrong, it's not that I don't want it to work, because it would be lovely if it did.

In most processors a type of relaxation oscillator is used to generate square waves to provide a clock signal. In such an oscillator, transistors are used. Let it be noted that high frequency begets higher voltages (voltage is the difference in potential--expressed as a line integral). As the frequency of electrons moving back and forth across the junction of a transistor increases, so does the potential for the electrons to skip right over the junction unexpectedly (read Peak Inverse Voltage).

This is why semiconductor oscillators are designed explicity for the specifications of the device to which they are to be used, and the paramaters of which governs the dilution/saturation of the silicon with impurities.

So, taking into account the voltage issue due to higher frequency, plus the intrinsic increase of voltage by the decrease in resistance due to decreasing the temperature, you would kill out the semiconductors in your processor. Not to mention the fact that your front-side-bus is no where near capabale of handling that kind of throughput. Nor can the ALUs or the pipeline cache.

Oh and just to point out a possible fallacy that might be found, when I'm talking about change in resistance, I'm talking about across the junction of a given semiconductor.

Any questions, class?

~zehly

(P.S. I'm only entertaining this subject because I can actually get away with talking about electrodynamics/thermodynamics -- something I love discussing.. par for the course in the life of a physicist, hehe. I'd be happy to talk to anyone about physics/electrodynamics/thermodynamics/etc. in PM if they need a deeper explanation, or can point you in the right direction, too.)

EchoSex · Nov 08, 2005, 10:24 PM // 22:24

Quote:

Originally Posted by zehly

Third party software, unfortunately, cannot change the thermodynamic properties of any open or closed system. Sorry, not happening. Don't get me wrong, it's not that I don't want it to work, because it would be lovely if it did.

In most processors a type of relaxation oscillator is used to generate square waves to provide a clock signal. In such an oscillator, transistors are used. Let it be noted that high frequency begets higher voltages (voltage is the difference in potential--expressed as a line integral). As the frequency of electrons moving back and forth across the junction of a transistor increases, so does the potential for the electrons to skip right over the junction unexpectedly (read Peak Inverse Voltage).

This is why semiconductor oscillators are designed explicity for the specifications of the device to which they are to be used, and the paramaters of which governs the dilution/saturation of the silicon with impurities.

So, taking into account the voltage issue due to higher frequency, plus the intrinsic increase of voltage by the decrease in resistance due to decreasing the temperature, you would kill out the semiconductors in your processor. Not to mention the fact that your front-side-bus is no where near capabale of handling that kind of throughput. Nor can the ALUs or the pipeline cache.

Oh and just to point out a possible fallacy that might be found, when I'm talking about change in resistance, I'm talking about across the junction of a given semiconductor.

Any questions, class?

~zehly

(P.S. I'm only entertaining this subject because I can actually get away with talking about electrodynamics/thermodynamics -- something I love discussing.. par for the course in the life of a physicist, hehe. I'd be happy to talk to anyone about physics/electrodynamics/thermodynamics/etc. in PM if they need a deeper explanation, or can point you in the right direction, too.)

*Raises hand* Yeah, why is it that you have no idea what you're talking about? Have you even TRIED to clock something? You know, it seems every day you claim to know more and more about something. Not trying to be an ass, but seriously - unless you've actually been out there and tried this kind of stuff, butt out.

Alias_X · Nov 08, 2005, 10:29 PM // 22:29

Wow Zehly, did you practice that? Well said, and if you can post the screen shots, I would like to see them, because it would like Zehly said be awesome if it were true, so please prove it is true!

zehly · Nov 08, 2005, 10:33 PM // 22:33

Quote:

Originally Posted by EchoSex

*Raises hand* Yeah, why is it that you have no idea what you're talking about? Have you even TRIED to clock something? You know, it seems every day you claim to know more and more about something. Not trying to be an ass, but seriously - unless you've actually been out there and tried this kind of stuff, butt out.

Yes, I have tried to overclock something. Yes, I am a graduate-level physicist. We've been over this, haven't we? Did you read the post above in which I stated that just today we tried to replicate the exact same thing that the OP did? I guess not.

Also, I don't post every day, and I don't make claims.

Avari Zi · Nov 08, 2005, 10:38 PM // 22:38

My knowledge of computing is limited, but even I have tried overclocking my Athlon. However, I found I could not overclock it by much due to the limits of the motherboard. My first attempt I had to reset the bios on my motherboard to even get my computer to start up again. How in the world did you get a Pentium III that high of a speeds and still have the system boot up regardless of heat or voltage? o.O

chris_nin00 · Nov 08, 2005, 10:40 PM // 22:40

*Leaning back on chair*
*Falls asleep, falls over*
Oh shiznit... I'm falling asleep in class.

Congrats Techie, I hope your overclocking doesn't hurt you!

Who cares if its real or not Zehly. Jeez. If he made it up then what does he get out of it?

*Raises hand*
I don't know...

I never overclock. I've taken it up a couple of times, read some books and magazines about it. But I never get started.

*falls asleep with cap on the face*

EchoSex · Nov 08, 2005, 10:41 PM // 22:41

Quote:

Originally Posted by zehly

Yes, I have tried to overclock something. Yes, I am a graduate-level physicist. We've been over this, haven't we? Did you read the post above in which I stated that just today we tried to replicate the exact same thing that the OP did? I guess not.

Also, I don't post every day, and I don't make claims.

You can claim all the education you want - the end result is what matters. The sum is greater than the whole - go out, over clock a PIII, THEN come back and say it can't be done. Until then, keep it to yourself.

zehly · Nov 08, 2005, 10:45 PM // 22:45

Quote:

Originally Posted by EchoSex

You can claim all the education you want - the end result is what matters. The sum is greater than the whole - go out, over clock a PIII, THEN come back and say it can't be done. Until then, keep it to yourself.

Can you not read? I overclocked a Pentium III today. You can overclock, yes. Can you use carbon dioxide solid to do it? No. You're a real jerk, EchoSex, I hope you know that.

/end

EchoSex · Nov 08, 2005, 10:47 PM // 22:47

Quote:

Originally Posted by zehly

Can you not read? I overclocked a Pentium III today. You can overclock, yes. Can you use carbon dioxide solid to do it? No. You're a real jerk, EchoSex, I hope you know that.

/end

Let's see the screenshots.

chris_nin00 · Nov 08, 2005, 10:48 PM // 22:48

/start
OooHH!!! Ohhh!!!
/pickme
Teach me how to overclock!!!
WEEEE!!!
I wanna light thing up!

EchoSex · Nov 08, 2005, 10:51 PM // 22:51

Quote:

Originally Posted by chris_nin00

/start
OooHH!!! Ohhh!!!
/pickme
Teach me how to overclock!!!
WEEEE!!!
I wanna light thing up!

Not something you want to get into headfirst, Chris. You can really end up damaging expensive hardware in the process, espcially if you decide to go clocking a $500 graphics card without proper knowlage and experiance. There's many guides on the net with links to software and vids that can help you. A good rule of thumb is to not overclock unless you're getting frames less than 30 with graphics all the way down on a game (for graphics cards, anyways.)