[SOLVED] Is passive cooling really more energy efficient than active cooling (Windows power management)?

Matiseli

Prominent
Oct 15, 2022
17
0
510
At first glance, passive cooling seems to be a more energy efficient option. But let's try to think about it more... I'll give an example:
I work in Adobe Premiere Pro. I have passive cooling set up in Windows. As a result, every operation in that application takes a little longer to complete. Lags (delays) add up, especially when you factor in video rendering.
Let's say that because of the passive cooling (or the wait associated with this form of cooling) it took me 100 minutes to make the video. In the case of active cooling, the same task took me 85 minutes. So the laptop ran for 15 minutes longer: it consumed electricity for 15 minutes longer.
So, is passive cooling really energy efficient? If my hypothesis is true, it seems that passive cooling is only relevant for tasks that are not demanding on resources: office work, light web browsing,...
Am I right?
thank you for answer
 
Solution
Hello thank you. Can you please explain the first paragraph better? I do not understand much. thank you very much

If automatic downclocking due to heat caused a task to take longer, it means that the parts were using less power. The isolated fact that there were 15 extra minutes is immaterial.

If there was no automatic downclocking due to heat, the task wouldn't take longer.

Passive cooling is always going to use less power, by definition. Power is heat and if you use active cooling to transfer heat, you're going to necessarily have to use some energy in the process. The laws of thermodynamics won't have it any other way. The reason we don't use passive cooling in high-powered PCs in most cases is because...
It also took less time because it used more power(because it was able to cool more power) ...
You have to measure the total power, of the whole system, used to finish the same workload and divide by the time it takes for both cases.

Also your situation isn't about active vs passive but about the amount of power YOUR passive cooling can handle compared to the power of YOUR active cooling.
If your passive cooling where just as powerful as your active cooling there wouldn't be a difference.
 
It also took less time because it used more power(because it was able to cool more power) ...
You have to measure the total power, of the whole system, used to finish the same workload and divide by the time it takes for both cases.
Hello thank you. Can you please explain the first paragraph better? I do not understand much. thank you very much
 
Hello thank you. Can you please explain the first paragraph better? I do not understand much. thank you very much

If automatic downclocking due to heat caused a task to take longer, it means that the parts were using less power. The isolated fact that there were 15 extra minutes is immaterial.

If there was no automatic downclocking due to heat, the task wouldn't take longer.

Passive cooling is always going to use less power, by definition. Power is heat and if you use active cooling to transfer heat, you're going to necessarily have to use some energy in the process. The laws of thermodynamics won't have it any other way. The reason we don't use passive cooling in high-powered PCs in most cases is because passive cooling is inferior at cooling, not that it's energy inefficient.
 
Solution
true passive cooling would use 0w, I have done this on multiple builds basically just utilizing 30+lbs of water and massive gauge copper wires to the heatsink/block etc. also is nice to add this water/wire combo to an existing build, thermal mass really cant be overstated how powerful it is, water has something like 300-1000x the amount of heat it can store compared to air so just imagine a 450lb barrel X that much air (also copper rods in the barrel!). should also note fans/pumps can easily pull over 100w in a powerful system under load, 3 gpu fans at max easily 20-30w+, 140mm or even bigger fans are by far best option per watt, possibly even a desk fan
 
Passive cooling is always going to use less power, by definition. Power is heat and if you use active cooling to transfer heat, you're going to necessarily have to use some energy in the process. The laws of thermodynamics won't have it any other way. The reason we don't use passive cooling in high-powered PCs in most cases is because passive cooling is inferior at cooling, not that it's energy inefficient.
That's not going to be true for modern chips and good coolers because today's manufacturing processes have such an enormous amount of leakage current. A hotter chip will have higher resistance and more leakage current so the same chip at the same voltage with the same workload will draw less power as the temperature goes down.

With a manual OC on a 7900X, Der8auer dropped power draw in Cinebench by delidding to drop temps. At ~90C the CPU drew a peak of ~219W and at ~71C peak power draw was ~187W. He used the same voltage and clock speeds for both runs.

If you use efficient fans like NF-A12x25s you're not going to be exceeding 1.7W per fan so with a big heatsink you could easily have lower overall CPU+FAN power draw than a huge passive cooler.
 
That's not going to be true for modern chips and good coolers because today's manufacturing processes have such an enormous amount of leakage current. A hotter chip will have higher resistance and more leakage current so the same chip at the same voltage with the same workload will draw less power as the temperature goes down.

With a manual OC on a 7900X, Der8auer dropped power draw in Cinebench by delidding to drop temps. At ~90C the CPU drew a peak of ~219W and at ~71C peak power draw was ~187W. He used the same voltage and clock speeds for both runs.

If you use efficient fans like NF-A12x25s you're not going to be exceeding 1.7W per fan so with a big heatsink you could easily have lower overall CPU+FAN power draw than a huge passive cooler.
This lower temp drawing less power effect is actually really large, when I make even a small i3 system I really prefer to use LM for this effect in particular pretty much. There is a very noticeable power draw diff of a cpu that gets to 70c max vs 50c 40c etc, the actual wattage difference in reality is pretty close to 1% lower per c from what I observe (huge diff). also yea, a fan drawing a few watts, ideally a 140mm will end up with a lower wattage system overall better due to that lower temp power draw cpu effect