PDA

View Full Version : Approx wattage of a 3.2E @ 4.3 GHz, 1.5Vcore?


AudiMan
09-17-04, 06:22 PM
Does anyone know of any CPU wattage calculators? I tried a few but they're not updated yet. I just need the approximate wattage of a 3.2E @ 4.3 GHz, at 1.5Vcore. D0 stepping. So stock would be 89W I believe.

hUMANbEATbOX
09-17-04, 08:16 PM
what's the stock vcore?

i put in 1.325 and got 153w from this calculator: http://newstuff.orcon.net.nz/wCalc.html

NeoGeo
09-19-04, 10:33 PM
I just need the approximate wattage of a 3.2E @ 4.3 GHz, at 1.5Vcore. D0 stepping. So stock would be 89W I believe.

Here is how to get the overclocked wattage with a good old calculator or a piece of paper and pen:

Stock: 3.2E
MHz= 3200 MHz
Volts= 1.40 V
Wattage (max according to intel spec sheet)= 89 watts

Overclocked Wattage = stock wattage X {(oc'd MHz/stock MHz) X [(oc'd volts/stock volts)^2])

Using 4300 MHz and 1.5 V would give you a maximum wattage of 137.29 watts.

Please note that if you are looking for real world results, Intel's maximum wattage rating is very hard to reach. So, look for typical wattage to be 65-75% of the maximum.

typical wattage at stock ~ 64 watts (taken @72% Intel's max)

typical wattage oc'd to 4300 ~ 99 watts (taken @72% Intel's max)

Any more questions?

Stratcat
09-19-04, 11:15 PM
Here is how to get the overclocked wattage with a good old calculator or a piece of paper and pen:

Stock: 3.2E
MHz= 3200 MHz
Volts= 1.40 V
Wattage (max according to intel spec sheet)= 89 watts

Overclocked Wattage = stock wattage X [(oc'd MHz/stock MHz) X (oc'd volts/stock volts)]

Using 4300 MHz and 1.5 V would give you a maximum wattage of 128.14 watts.

Please note that if you are looking for real world results, Intel's maximum wattage rating is very hard to reach. So, look for typical wattage to be 65-75% of the maximum.

typical wattage at stock ~ 64 watts

typical wattage oc'd to 4300 ~ 92 watts

Any more questions?
Sorry NeoGeo, but I've gotta' respectfully disagree, in part:

Power dissipation scales linearly with clockspeed, BUT, scales as the square of Voltage. So your equation's delta Vcore component should look something more like this: [(oc'd volts/stock volts)^2]

Otherwise, it looks fine, to me.

Putting it all together, it would then become: Overclocked Wattage = stock wattage X [(oc'd MHz/stock MHz) X (oc'd volts/stock volts)^2]

Nice job, and a very minor error, IMO.

Not sure where I remember the linear clockspeed scaling, and voltage squaring, but it has popped up a few times over the years. Maybe an html or Usenet Google will come up with somethin'. I'm pretty sure I'm right, on the squaring of the voltage, though.

I'll let you guys work out the revised #'s.

Strat

AudiMan
09-19-04, 11:33 PM
I'm just interested to know the heat output of a highly overclocked prescott because it's good to know when you're in the market for a phase change system. 128 watts is peanuts.

meionm
09-19-04, 11:37 PM
Here is how to get the overclocked wattage with a good old calculator or a piece of paper and pen:

Stock: 3.2E
MHz= 3200 MHz
Volts= 1.40 V
Wattage (max according to intel spec sheet)= 89 watts

Overclocked Wattage = stock wattage X [(oc'd MHz/stock MHz) X (oc'd volts/stock volts)]

Using 4300 MHz and 1.5 V would give you a maximum wattage of 128.14 watts.

Please note that if you are looking for real world results, Intel's maximum wattage rating is very hard to reach. So, look for typical wattage to be 65-75% of the maximum.

typical wattage at stock ~ 64 watts

typical wattage oc'd to 4300 ~ 92 watts

Any more questions?


4300 92watts, you are so off you could double it, at 4300 with 1.5 at least 160watts that what I would say, some prescott cpu will run cooler, but what I seen it takes a prommy to 0c at 4400 with 1515v, and prommy should take about 220watts of heat before reaching zero

according to sisandra I am reaching 150 watts and I am not even running prescott

NeoGeo
09-19-04, 11:59 PM
So your equation's delta Vcore component should look something more like this: [(oc'd volts/stock volts)^2]

No need to google it, you're right Strat I left out the squaring of the voltage difference. I will edit my post to reflect what you have corrected.

NeoGeo
09-20-04, 12:17 AM
4300 92watts, you are so off you could double it, at 4300 with 1.5 at least 160watts that what I would say, some prescott cpu will run cooler, but what I seen it takes a prommy to 0c at 4400 with 1515v, and prommy should take about 220watts of heat before reaching zero

according to sisandra I am reaching 150 watts and I am not even running prescott

You're just pulling numbers from thin air aren't you. How do you conclude 160 watts?

Using a tried and true wattage equation with 4.4 GHz and 1.515 V would only yield a "theoretical maximum" of only 143.3 watts. And, that maximum could not be reached by ordinary means even running dual whatevers (f@h, prime95, etc) and 3dmark and Sandra and Super Pi...

From what I have seen, both Vapochill PE and LS along with a few air coolers, you could reach at most approximately 80% of Intel's maximum thermal guideline. And Sandra can be quite far from accurate when it comes to measuring a few things such as temperature, voltage, and wattage.

As to the 0 deg C temps read from a Prommie, temps in real world conditions have many factors other than the wattage output of a cpu, in particular the ambient temperature, the thermal paste used, the thermal probe and where the thermal probe is applied, the thermal conductivity of the cooling head, etc...