Permit me to disassemble your orginal remarks once again in an attempt to grasp your point(s).
is it really that necessary to have such an elaborate cooling unit?
I know how overclocking works, and if you knew, you would recognize that the cooler in that pic is kinda extravagant
Do you really know how it works?
First of all when you build something in the hopes of acheiving new limits, do you build it so that it will be "just enough", or do you build it so it with some (calculated) leeway? The limits of the particular p4 would not be know until it was actually tested, so based on estimated heat output among other factors he build a system using his technical knowledge and experience in an attempt to see how far he could push technology. He didn't know before what limit would be acheived. All cpus have limits..but it can't be known until pushed.
Elaborate? By who standards? Yours?
Have you ever tried serious o/cing? Fugger is deemed an o/cing expert by his peers and I have been following his exploits for a year now personally so I totally agree. Its not just his accomplishments but the manner in which he acheives it. If thats what was needed then how can be deemed "elaborate"? Are you telling me you could acheive 6+Gig on significantly less / "less elaborate" a setup? He worked within his restraints. The most interesting fact about his system is that its sustainable...unlike the crazy LN2 Japanese. If is so easy to reach so high an o/c on a "less elaborate" system then why so few have reached anywhere near this level?
If overclocking a 1 gig cpu can boil liquid rocket (as in space shuttle, as in NASA) coolant, what makes this 6 gig cpu so much different!!!!!!
Lets just assume your making perfect sense for a bit. The two conclusions I have formed your "facts" thus far are:
#1 1gig boils liquid coolant so ]1gig = boil coolant = boil coolant faster.
#2 Your comparing *liquid* rocket coolant + space shuttle temp vs gas + cpu temps (but not even in same kind of environment as in rocket)
Is your point that he should have "simply" used liquid rocket coolant?
And what does it boiling so "easily" have to do with making it a better choice as coolant as you implying?
Or, how does it make less elaborate the cooling process required to attain such an o/c? Simply submerge the p4 in LRC according to your logic?
It would evaporate fast, like isoprophyl alcohol, but it would not boil! The liquid rocket coolant only began to boil when the mobo containing the active cpu was placed in it. This is rocket coolant, the stuff designed to keep ROCKET engines cool. that is impressive.
What exactly is impressive? The LRC or the cpu or boiling?
#1 the coolant in Fugger system is NOT exposed to AIR nor can it "boil" as a liquid can.
#2 In a shuttle, liquid rocket coolant IS NOT EXPOSED TO AIR.
Was the mobo sealed air tight? No so whats your point? Again it seems like your saying all he had to do is use LRC in his system instead of all that...but then EVEN IF that did make sense...consider below.
Lets analyse this "elaborate" system a bit:
What makes it "elaborate"...im guessing the powerful compressor and massive radiators/condenser.
The purpose of the Compressor is to condense the gas to liquid form...and that will take a MASSIVE amount of "sucking" energy/power.
His massive radiators/condensers...Well without them where would all the massive heat be dissipated...so even if LRC was "better"...you still need all that "elaborate" stuff to dissipate away the heat from the cpu/system.
In other word we really do need all this "crap".
Lets say using LRC aka liquid was feasible to him...How exactly would this work (scientific wise) to allow so low temps?
Now before I go any further...You obviously have not considered the difference between keeping somthing cool...and keeping something cool below a certain temperature.
Its relatively simple keeping something cool as in around room temperature...but its exponentially more difficult to keep below room temperature.
Rockets dont need to be cooled below room temp..in fact they can and do withstand way above room temp as norm.
Cpus on other hand when being pushed to limits need especially sub-zero temps...constant sub-zero temps.
The "cooling" power of a coolant is determined by how much heat it can absorb in unit time. In chemical terms its specific latency heat AND NOT BOILING POINT.
VAPOR PHASE COOLING (same as how your fridge works) - The coolant aka gas is converted to liquid form by compressor. This lowers temperatures DRASTICALLY since it takes a HUGE amount of energy to convert form ESPECIALLY since its against its natural state (note change of form can be natural or forced a point which will be explain lower).
Now what Fugger did was 3 stage..instead of one like in fridge...At each stage he further enhances cooling by cooling the already cooled coolant. Thats how he gets so low temps.
* SEE CHEM 101 below for more details on the logic behind vapor phase cooling aka the cooling Fugger used.
3.6GHz @ 80-105W, 6GHz would be wha..... close to 200W??? plus cooling pumps etc.
a 3.6 Prescott gives out 115W+ at default voltage..@6G...I would say ~300w especially given his voltage increase.
2x 3ghz does not = 2x performance . He has bottlenecks in memory bandwidth , bus speed , cache and latencies all round . Its not really worth the money for just the performance, but its a kewl hobbie
We all know that. But thats with all else equal. And in this case all else IS NOT EQUAL.
There will always be bottlenecks but you obviously dont give him ANY credit for his amazing reduction in bottlenecks...he just didnt acheive 6g+ he o/ced the hell out of everything else. Everything he used was cutting edge...and he pushed each and every component till it bled. Even if he didnt reach so high an o/c just pushing his other components so high is a feat in itself.
LET ME SPOON FEED YOU NOW:
CPU FSB = 450!!!! MEM FSB = 350!!!
Let me help you out some more...pc3200 is equivalent to: CPU FSB = 200, MEM FSB = 200.
Thats HEAVENLY bandwidth there! But you failed to take that into consideration. Instead you just quoted a typical for the mneub disclaimer. Which otherwise would'nt be so out of place had it been a normal situation.
CONCLUSION ]] Your talk of bottleneck is POINTLESS since his memory, bus, etc are way above everything previously known to mankind.
CHEM 101:
In transforming from a liquid to a gas heat is absorbed...and thus it "cools". That said, all coolants have different properties and different cooling efficiency based on situation. Namely the value known as specific heat capacity. Although its not simple as just that one value if you chemically inclined. For example suppose we used say a substance known as ChemicalX in our cooling system. Now say it become gas at 1 million degrees celcius aka boiling point is 1M degrees Celcius. Now you can just hear
camre bawling down the place proclaiming this to be the ULTIMATE coolant! But my dear camre that be incorrect it so happens ChemicalX can't absorb heat for nothing...it cant even "flow" for that matter! The point (albeit exaggerated) is BOILING POINT HAS LITTLE OR NOTHING TO DO WITH COOLING POWER. All that matters is how effectively it can absorb and dissipate energy == Specific heat, Latent Heat and Vapor Pressure, etc to be absolutely anal. Of course if you using in environment in which temps gets so high that caused a change in form in the coolant WHICH YOU DONT WANT, THEN it becomes concern...but for what we discussing here ITS TOTALLY IRRELEVANT.
Which brings me to the point why gas coolant will always be greater than liquid or other coolant all else equal. Consider in turn:
Liquid:
liquid -] gas -] liquid (only one stage we going against the natural state of substance)
Vapor Phase:
gas -] liquid -] gas -] liquid (TWICE we going against the natural state of substance)
In plain engrish, Those two forcings allow for much enhanced heat absorption powers than the one...even ignoring specific latent heat, etc values of the coolants involved.
When the dude put the mobo in the coolant and started the pc, it ran for a couple minutes and began to BOIL the ROCKET COOLANT
Umm a cpu can't exceed 100 degrees period. It would die way before that so how could it boil NASA stuff unless LRC was designed with such a low boiling point on purpose. Which would make some sense since boiling faster mean faster sucking of energy (during state change)...but that still only be useful for QUICK temp changes...like only rockets can do. A cpu has extremely gradual temp increase by comparison.
In other words I can't possible see what benefit LRC could have. Your logic is flawed any way you take it.
As if thats not enough, how do you know its LRC used for sure anyway? That guy could have used anything. And what exactly is LRC? Im sure there are many kinds which varying compositions...further adding variables to the argument.
