Vapochill LS and Mach ST/GT test part1

[JCviggen]

Originally posted by phobix
@pkrew

Use the following tool to calculate overclocked watttage

http://www.benchtest.com/calc.html

One thing to keep in mind though, generally speaking you never get even close to the theoretical wattages, as I noticed. Maybe 70% of that....

If the wattage calculator says 200W maybe you will reach 140-150 actual when doing prime and CPU burn in ACTUAL output.

Its often much less heat being put out that you think

[eva2000]

when sandra burnin cpu runs it shows a cpu wattage... how close/accurate would that be ?

[JCviggen]

Originally posted by eva2000
when sandra burnin cpu runs it shows a cpu wattage... how close/accurate would that be ?

not close at all. it just takes a guess based on speed, voltage and the core type. Its probably about the same as when you calculate it on the above formula. Its a theoretical maximum.

[Formann]

Heres a calc that diplays normal output aswell max teroretical output.

http://www.ocshoot.no/cpuheat.htm

Its in norwegian, but im sure you´ll understand. Just choose CPU on the top, then fill inn overclocked cpu-speed and voltage. Then press "Kalkuler"

It hasn´t been updated for a while, but I´ll se what I can do. It lacks the A64´s and the preshots.

IE.
Barton - 2870Mhz @ 1.97v - Normal output 134.4w - Max Teoretical output 152.7w

P4 2.5 - 4150Mhz @ 1.9v - Normal output 183.1 - Max Teoretical output 208w Damn.. then imagine the prescott @ 4.5Ghz.

[SAE]

Originally posted by Formann
IE.
Barton - 2870Mhz @ 1.97v - Normal output 134.4w - Max Teoretical output 152.7w

P4 2.5 - 4150Mhz @ 1.9v - Normal output 183.1 - Max Teoretical output 208w Damn.. then imagine the prescott @ 4.5Ghz.

That's more realistic

[ricjax99]

Originally posted by Formann
Heres a calc that diplays normal output aswell max teroretical output.

http://www.ocshoot.no/cpuheat.htm

Its in norwegian, but im sure you´ll understand. Just choose CPU on the top, then fill inn overclocked cpu-speed and voltage. Then press "Kalkuler"

It hasn´t been updated for a while, but I´ll se what I can do. It lacks the A64´s and the preshots.

IE.
Barton - 2870Mhz @ 1.97v - Normal output 134.4w - Max Teoretical output 152.7w

P4 2.5 - 4150Mhz @ 1.9v - Normal output 183.1 - Max Teoretical output 208w Damn.. then imagine the prescott @ 4.5Ghz.

On that site select 2800+ instead of 2500+ and the wattage same overclock becomes lower

[Ragnarok]

coz the calculation also takes into account the default speed..

[ricjax99]

Originally posted by Ragnarok
coz the calculation also takes into account the default speed..

Eh, so if it follows true:

A Barton 2500+ running at 2400 MHz has a higher wattage than.....

A Barton 2800+ running at 2400 MHz.

So does it take into account the FSB increase it would take to get a 2500+/2800+ to 2400 MHz with its default multiplier?

As 2500+ would need 218x11 and the 2800+ would need only 192x12.5 hmm

Seems a bit odd thats all.

[Bridgeboy]

Originally posted by ricjax99
Eh, so if it follows true:

A Barton 2500+ running at 2400 MHz has a higher wattage than.....

A Barton 2800+ running at 2400 MHz.

I'm certainly no expert, but I have read that different versions of the same core architecture of a particular processor type are possible simply because the same processor rated at the higher speed architecture was fabricated either to higher tolerances, or, just came out of the fabrication process (with same tolerances used for 2500+ models) with better quality and is capable of higher clock speeds while still remaining stable. Therefore, being that the Barton 2800+ is rated higher than the 2500+ yet has the same core architecture, it probably means that the 2800+ can run cooler than the 2500+ at any given clock speed because it is simply a higher quality fabrication of the same chip.

This is just my own logic based on things I have read, however it may not be completely true in actuality. But it makes sense to me.

[zabomb4163]

Originally posted by JCviggen
not close at all. it just takes a guess based on speed, voltage and the core type. Its probably about the same as when you calculate it on the above formula. Its a theoretical maximum.

No offence JC, but i believe you may have confused the intel thermal guideline with the amd thermal guideline.

they are not equal. AMD list the theoretical maximum heat output of their processors whereas intel list the normal heatload of the processor and refuses to list the theoretical max heat output #'s.


http://www.amd.com/us-en/assets/cont...docs/20092.pdf

Pmax is the maximum power consumption (in watts) of the

ftp://download.intel.com/design/inta...d/27371601.pdf

[JCviggen]

Originally posted by zabomb4163
No offence JC, but i believe you may have confused the intel thermal guideline with the amd thermal guideline.

Not at all

It is my experience that the actual output of the CPUs tends to be below what you would expect, wether its Intel or AMD. The numbers are always based on worse case scenarios it seems. And then some.

And as said earlier, some of these formulas dont add up. A 2500+ or 2800+ Barton have the same output at the same settings, because they are the same core (at least now they are) unlike what some formulas suggest.

Going by math, you could get a current CPU to deliver 200W. Well, from what i have seen, you realistically cant.

[Formann]

The thermal output should indeed be the same on a XP2500+ and a XP2800+ when running the same speed and vcore.

I cant remember where we got the formulas, but i will look in to it. The whole ting is a beta, but the calculations are reasonable, and shouldn´t be very far off.

[phobix]

Originally posted by JCviggen
[B
you could get a current CPU to deliver 200W. Well, from what i have seen, you realistically cant. [/B]

Ah but wait... Introducing Prescott, your all purpose furnace

[zabomb4163]

Originally posted by JCviggen
Not at all

A 2500+ or 2800+ Barton have the same output at the same settings, because they are the same core (at least now they are) unlike what some formulas suggest.

http://www.amd.com/us-en/assets/cont...docs/26237.PDF

For 2500+ @ 3200+ (2200/1833) * 53.7 W = 64 W typical
* 68.3 W = 81.97490453

according to amd tech docs

3200+ Typical heat = 60.4 W
3200+ thermal max = 76.8 W

so 4 watts off on the typical and 6 on the max.

[CrashOv3r1De]

Great results! Thanks a lot for sharing. Im not in the boat for a new phase change cooler yet (still using my new mk2) but these results are really great for users who are planning on purchasing their 1st phase change unit. Looks like the noise/price will be thew biggest factor in the decisions.

JC the mk2 was estimated for $980usd why are people selling it for $1200+?

[aoc007]

So the Vapo LS is the noisiest of the three?

[`schr0et]

Mirror mirror on the wall, who's the noisiest of them all?

I dunno but if the LS can hold decent temps @ 200W+ (if thats even possible) then it seems like a decent investment for the future CPUs.

But for me at this moment... nah

`s

[petervandamned]

in part 2 i will add some avi's with sound

[interman]

any idea on the eta?

[petervandamned]

just waiting on a few parts for the testbank

[Çhrist0ph]

ok, i almost started a thread for my question, but why not ask it here?

if the vapochill lightspeed is charged/tuned/optimized for lower temps @ 240+ watt heatload, wouldnt pinching its cap tube help it getter better temps under todays loads(100-150watt,etc?)

if i know anything about refridgeration, isnt a system tuned for a high heatload have a lot of liquid refridgerant flooding the evap, waiting to be heated up/evapoarated by a huge heatload, but only finds a small heat load so most of the liquid refirdgerant just goes back barely evaporated? therefore, temps arent as good as they should be ?

[Ragnarok]

look, asetek/nventiv are smart enough to take care of the liquid return issue..

otherwise under CPU idle, u r gonna get heaps of liquid return..

the vapo LS is tuned for a higher load thing hence imo is a load of bull..

[comment]

Originally posted by Ragnarok
look, asetek/nventiv are smart enough to take care of the liquid return issue..

otherwise under CPU idle, u r gonna get heaps of liquid return..

If the liquid did return, would the result be seething coming from the compressor?

[Ragnarok]

well try compressing a liquid, all u gonna get is a ed compressor

[comment]

Originally posted by Ragnarok
well try compressing a liquid, all u gonna get is a ed compressor

All right then. But what would seething from the compressor indicate?