Computer cooling off period

[Rocketdog2112]

I always tell a buyer of a gaming computer they are buying from me that they should allow sufficient time to allow the desktop to cool down after intensive gaming. I get some very strange and surprizing looks sometimes.

 

That got me wondering what everyone here does. I myself always allow the GPU's time to cool down as they are always the hottest item in the case. I just don't like the thought of the GPU's and other components becoming heat sinks with no air circulation.

 

My GPUs (EVGA 1080 FTW 2 in SLI) are usually in the mid 30's before I shutdown the system.

[matt0725]

nothing will cool the devices faster than turning them off, therefore stopping heat generation. In the time it takes you to quit a game and your pc to shut down, core temps on the GPU and CPU go back close to idle. the heatsinks may still be warm, but theyll do what theyre designed to do, radiate heat out.

[TheBean]

I will sometimes just let fans go to 100%. shutting it down will turn off the fans so I dont turn it off

[Holmes108]

Interesting, I've never heard of cooling it off. 

[Bitter]

Seeing as there's not much mass in the hot parts of a system I don't think there's a lot of issue with thermal spike with sudden loss of fan once the load is gone.

[Stahlmann]

Just as a PC doesn't need time to get to operating temperature you don't need to let it cool down before shutting it off. There is no oil cycle with an optimal operating temperature in your gaming PC that you have to think about. A PC isn't a car.

[Heliian]

1 hour ago, Stahlmann said:

A PC isn't a car.

No, it is not, but a lot of things aren't cars. 

 

Can't hurt by letting everything spool down before shutting it off.

[BetteBalterZen]

Never made sure to let my PC cool off and never will 

Also, the instant I close my game, CPU and GPU drops to 30/40C anyways. 

[Stahlmann]

1 hour ago, Heliian said:

No, it is not, but a lot of things aren't cars. 

 

Can't hurt by letting everything spool down before shutting it off.

Sure it doesn't hurt. But it doesn't benefit you either. It's one of these things you really do not have to worry about - at all.

[LapsedMemory]

Wait...you people shut your PC off?

[Lurking]

Just the few seconds with no load when you exit the game to going to shutdown will remove any heat left.

 

This isn't a 1970's oil cooled turbocharger. You give your clients bad advice.

[Middcore]

25 minutes ago, Lurking said:

Just the few seconds with no load when you exit the game to going to shutdown will remove any heat left.

 

This isn't a 1970's oil cooled turbocharger. You give your clients bad advice.

Well, it isn't bad advice in the sense that it will hurt anything, but it's basically superstitious "folklore" that has no benefit and perpetuates needless anxiety about how fragile PC components are, and there's already enough of that in this hobby.

[PC Master Blaster]

Maybe 25 years ago when computer cases only had one 80mm fan but haven't had that thought in awhile.

[Lurking]

39 minutes ago, Middcore said:

Well, it isn't bad advice in the sense that it will hurt anything, but it's basically superstitious "folklore" that has no benefit and perpetuates needless anxiety about how fragile PC components are, and there's already enough of that in this hobby.

Giving bad advice always  is bad. Apparently OP's clients noticed he is full of it (based on their reaction). So they also will dismiss other advice OP gave, even if that actually was good advice. and someone who actually believes this shut-down procedure and follows through, will do less other actually useful maintenance. Someone giving bad advice destroys the trust a consumer should have in a sales person. So there are negative consequences somewhere. 

 

For example, if OP told them about this shut down procedure and to also clean the inside from dust and hair every once a while, they will dismiss both parts. Even if the part about keeping the PC clean is good advice. 

[starsmine]

If a normal person said this, I would politely and calmly educate them.

Considering you found yourself in the authority position to sell PCs. I will ridicule you for such a dumb take. This was awful and we are all dumber for reading this. Do not make people unnecessarily anxious or paranoid because of your factually incorrect based on nothing statements like that OP.

[Bitter]

Ok maybe 30 years ago this was a thing when people were overclocking Pentium 2's and there was a slight possibility that latent heat could continue to rise in the case post shut down and degrade mechanical hard drives and motherboard capacitors. I remember my friend who was into overclocking talking about this, he read it on an email message board But now it's really not a concern, heat tolerance, case airflow, etc are all so much better. The only time I might consider this to be a thing would be if you had a water cooling loop that was running hot and was heat soaked, so a lot of thermal mass and maybe some thermal inertia from heavy water blocks, maybe letting that sit hot would be bad for something, maybe? Probably not, if you water loop is running THAT hot you need to fix stuff, not let it run a cool down lap.

[Lurking]

21 minutes ago, Bitter said:

Ok maybe 30 years ago this was a thing when people were overclocking Pentium 2's and there was a slight possibility that latent heat could continue to rise in the case post shut down and degrade mechanical hard drives and motherboard capacitors. I remember my friend who was into overclocking talking about this, he read it on an email message board But now it's really not a concern, heat tolerance, case airflow, etc are all so much better. The only time I might consider this to be a thing would be if you had a water cooling loop that was running hot and was heat soaked, so a lot of thermal mass and maybe some thermal inertia from heavy water blocks, maybe letting that sit hot would be bad for something, maybe? Probably not, if you water loop is running THAT hot you need to fix stuff, not let it run a cool down lap.

That is a trustworthy and scientific source....good reason to still follow the same advice in 2022

 

Pentium PCs, even overclocked, had so much less power consumption than modern gaming PCs. It was even less an issue than now. And once you stop your application (i.e. running Prime or whatever one did with a Pentium) the load immediately stopped while the fans were running. it took at least 5 seconds with fan still running to shut down, and then windows actually shutting down some more seconds with fans running and not much heat produced. So the chips at that point already were at a lower temperature than when running under load. 

 

You can monitor CPU temp and run a load (i.e. Prime). Once you turn off the load, you see CPU temp falling off a cliff within 1 second. 

 

And a PC doesn't have a latent load. The CPU/GPU have a specific balanced temperature when running at high load, and every item in the cooling chain (spreader paste, sink etc.) are colder. That is how heat transfers from hot to cold. Even if you unplug the PC without shutting down (you stop the fan in middle of hot workload) the chips will just cool down, not heat up more. This isn't a nuclear reactor where there still is some reaction happening years after. 

 

This isn't an old turbo charger where the turbine actually still runs and then slows down for a few seconds (by producing friction heat) after the motor stops. And even that was only an issue if one ran the engine at high rpm and then suddenly turned it off (unlikely real world scenario). and it was only an issue with oil-cooled turbos. Modern turbos are water-cooled (more thermal mass). So even that isn't an issue and none of the modern turbo cars need a specific procedure. But I bet, some car sales person still tells people to wait 5 minutes idling before tuning off the engine. 

[FuzzyLogic]

I've never heard of cooling down a PC first, and I'll never do it. Why? Because cooling down isn't necessary for electronics.

[Bitter]

1 hour ago, Lurking said:

That is a trustworthy and scientific source....good reason to still follow the same advice in 2022

 

Pentium PCs, even overclocked, had so much less power consumption than modern gaming PCs. It was even less an issue than now. And once you stop your application (i.e. running Prime or whatever one did with a Pentium) the load immediately stopped while the fans were running. it took at least 5 seconds with fan still running to shut down, and then windows actually shutting down some more seconds with fans running and not much heat produced. So the chips at that point already were at a lower temperature than when running under load. 

 

You can monitor CPU temp and run a load (i.e. Prime). Once you turn off the load, you see CPU temp falling off a cliff within 1 second. 

 

And a PC doesn't have a latent load. The CPU/GPU have a specific balanced temperature when running at high load, and every item in the cooling chain (spreader paste, sink etc.) are colder. That is how heat transfers from hot to cold. Even if you unplug the PC without shutting down (you stop the fan in middle of hot workload) the chips will just cool down, not heat up more. This isn't a nuclear reactor where there still is some reaction happening years after. 

 

This isn't an old turbo charger where the turbine actually still runs and then slows down for a few seconds (by producing friction heat) after the motor stops. And even that was only an issue if one ran the engine at high rpm and then suddenly turned it off (unlikely real world scenario). and it was only an issue with oil-cooled turbos. Modern turbos are water-cooled (more thermal mass). So even that isn't an issue and none of the modern turbo cars need a specific procedure. But I bet, some car sales person still tells people to wait 5 minutes idling before tuning off the engine. 

I always had my doubts, but if you remember the old hex pin grid heatsinks that were milled out of copper and weighed like 3lbs then I could see how the idea came to be.

[cmndr]

This shouldn't matter in any meaningful way. Temps drop within seconds of halting the heavy activity and heatsinks and water coolers modulate temperature shifts. 

While an argument could be made for letting things cool off when it comes to extending device lifespan, the difference is so small it doesn't matter. If you're super worried, get better cooling. 

[LogicalDrm]

*** Thread locked ***

 

Not seeing this going to any brighter direction.