CPU with higher TDP - Impact on needed fan speed?

Cooling Processors quietly

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
spookmineer
Patron of SPCR
Posts: 749
Joined: Sat Nov 11, 2006 6:02 pm

CPU with higher TDP - Impact on needed fan speed?

Post by spookmineer » Fri Sep 24, 2010 2:20 pm

The description is a bit cryptic...

Someone has a 65W CPU and cools it using a heatsink with a fan at 600 RPM.
What would be the impact on the needed fan speed if he went for a 95W CPU?
Or even a 125W CPU?

I try to imagine what the impact would be but it's hard.
Most of the time, the CPU will idle or have a low load, so the impact might be negligible in most cases.

But at full load, could it be as simple as: going from a 65W to 125W CPU, the needed fan speed doubles (if you want to get the same temperature)?

Does anyone have an idea? I've got a 65W CPU atm, but I'm thinking about getting a 4 or 6 core, but they are "hot".

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Fri Sep 24, 2010 3:27 pm

This is far too dynamic / complex a situation for anyone to answer with any hope of rational precision. Even if you had every detail (heatsink type, surface area, test specs, ambient temp, fan model, etc) your question cannot be answered accurately without empirical testing. ie -- just trying it is the only way.

spookmineer
Patron of SPCR
Posts: 749
Joined: Sat Nov 11, 2006 6:02 pm

Post by spookmineer » Fri Sep 24, 2010 3:53 pm

I was afraid you were going to say that... ;)

Was hoping for some of the experiences, maybe someone here has changed the CPU in an otherwise unchanged system and remembers/logged the changes in temperature, or rather how much faster the fan had to go to keep things in the same temperature range.

Some reviews state the temperature change when overclocking a certain CPU, and although it's not the same for every setup, people have an indication of the impact on temperature when overclocking.

In the same way, it would be informative if someone can share their experience on fan RPM when overclocking, or even changing the CPU. No exact figures, just an indication...

mkk
Posts: 687
Joined: Sun Sep 05, 2004 1:51 pm
Location: Gefle, Sweden
Contact:

Post by mkk » Sat Sep 25, 2010 1:56 am

While their TDP values aren't too specific about how much heat they will really produce I'd say avoid the 125W ones as a basic rule if you also want to do some overclocking on top of that. Using a case with good airflow and a big CPU cooler I'm looking at the AMD X6 1055T, but I would not aim to overclock. Silence is king and idle power usage is another factor that is getting to me. Overclocking to me has also always been about taking some inexpensive chip and doubling its monetary value. Taking a mid range CPU and overclock some 20% is not giving me much. ;)

tim851
Posts: 543
Joined: Wed Aug 13, 2008 11:45 am
Location: 128.0.0.1

Post by tim851 » Sat Sep 25, 2010 2:13 am

spookmineer wrote:No exact figures, just an indication...
Well, most likely, temps will go up. Vague enough?

At the very least, you need to post what CPU you are using now, which one you plan to purchase, the heatsink you're using, the fan, the fan's present RPM and the resulting temps and what kind of temps you're hoping for with the new setup.

Then maybe somebody with a similar setup and experience can report back.

Das_Saunamies
*Lifetime Patron*
Posts: 2000
Joined: Tue May 15, 2007 1:39 am
Location: Finland

Post by Das_Saunamies » Sat Sep 25, 2010 2:38 am

My 2 cents from a personal account:

Used the Noctua NH-U12P (old model) on my E8400 (65 W, dualcore). Idle was <40'C, load never went over 50'C. Bought an i5-760 (95 W, quadcore) and a mounting kit for the cooler. With the same cooler at the same RPM (900-1000 RPM with the ULNA), idle is still <40'C (38'C typical) but load goes up to 56'C (Orthos on two cores, Bad Company 2).

This is all in the same P182 case as the previous E8400 rig.

600 RPM is much lower than what I have (because I have to use the ULNA instead of a proper controller), so depending on the cooler and ventilation, I would think it wise to ramp it up for a 125 W CPU. 95 W you might still get away with just temperature rise, but if the rest of the rig is as low-flow as the CPU cooler, would suggest a hike in that case too, just to guarantee component life.

spookmineer
Patron of SPCR
Posts: 749
Joined: Sat Nov 11, 2006 6:02 pm

Post by spookmineer » Sun Sep 26, 2010 12:36 pm

Das_Saunamies wrote:Used the Noctua NH-U12P (old model) on my E8400 (65 W, dualcore). Idle was <40'C, load never went over 50'C. Bought an i5-760 (95 W, quadcore) and a mounting kit for the cooler. With the same cooler at the same RPM (900-1000 RPM with the ULNA), idle is still <40'C (38'C typical) but load goes up to 56'C (Orthos on two cores, Bad Company 2).
That's great news, only an increase of less than 10 degrees for the extra 30W.

In the end, MikeC is right and I just have to try.

CA_Steve
Moderator
Posts: 7651
Joined: Thu Oct 06, 2005 4:36 am
Location: St. Louis, MO

Post by CA_Steve » Sun Sep 26, 2010 2:44 pm

Another reference point for you...nothing absolute, but can give a feel for your cooling potential. Run Orthos or Prime95, etc with the current cpu and see how long it takes for the load temps to stabilize (ignoring the slow internal case rise effects). Stop the app and see how quickly the temps drop to or near idle temps. If it only takes a couple of seconds to stabilze at load and a couple of seconds to get near idle, then your cooler/fan are able to easily remove heat from the cpu. If you have a long tail from load to idle, then your cooling solution is struggling.

Post Reply