How dangerous is CPU overheating

[Feconi]

I understand that when the processor overheats, the degradation rate of the crystal increases (despite the fact that the degradation time of the crystal without overclocking is calculated for decades and the computer is updated faster than the processor degrades) and lowers the operating frequency (as I understand it, games mainly load the graphics card, so a slight decrease in the operating frequency of the processor will not be noticeable).

 

Can overheating of a modern processor be really physically dangerous for a computer if other components, such as mosfets, have a normal temperature?

[filpo]

Just now, Feconi said:

Can overheating of a modern processor be really physically dangerous for a computer if other components, such as mosfets, have a normal temperature?

depends how far they overheat. Most cpus have a temperature limit nowadays but it could also degrade the fins in a radiator or the fins in a heatsink with erratic changes in temperature

[Somerandomtechyboi]

13 minutes ago, filpo said:

degrade the fins in a radiator or the fins in a heatsink with erratic changes in temperature

How tf

like performance degradation of a heatsink? I geniunely dont see how thats possible if the metal is simply slightly warped

 

 

As for silicon degradation it is linked to both voltage and heat, you can run 115c just fine if you got a low enough vcore (like <1v) but if you run a high vcore (>1.5v) you my start to see degradation even at a mere 80c, the former is a somewhat educated guess so take that with a grain of salt

 

i reccomend following the manufacturers tjmax but if you wanna determine for yourself you can also reference other stuff, like rx 6600 having a tjmax of 110c at whatever stocm volt those run

[filpo]

1 minute ago, Somerandomtechyboi said:

How tf

like performance degradation of a heatsink? I geniunely dont see how thats possible if the metal is simply slightly warped

in my head it seemed to work 

 

Then again my head thinks that linus can go a whole vid without dropping anything

[Feconi]

5 minutes ago, Somerandomtechyboi said:

How tf

like performance degradation of a heatsink? I geniunely dont see how thats possible if the metal is simply slightly warped

 

 

As for silicon degradation it is linked to both voltage and heat, you can run 115c just fine if you got a low enough vcore (like <1v) but if you run a high vcore (>1.5v) you my start to see degradation even at a mere 80c, the former is a somewhat educated guess so take that with a grain of salt

 

i reccomend following the manufacturers tjmax but if you wanna determine for yourself you can also reference other stuff, like rx 6600 having a tjmax of 110c at whatever stocm volt those run

I'm talking about cpu overheating in stock settings, without changing the voltage and tjmax

[Somerandomtechyboi]

Just now, Feconi said:

I'm talking about cpu overheating in stock settings, without changing the voltage and tjmax

At stock settings who tf cares

Its fine upto tjmax so as long as it isnt throttling then its fine to run all the way to tjmax

[Feconi]

15 minutes ago, Somerandomtechyboi said:

At stock settings who tf cares

Its fine upto tjmax so as long as it isnt throttling then its fine to run all the way to tjmax

Lol I care  If three-section water cooling can keep the fervor of the i9-12900K in games around 65-75 degrees, then in the render the temperature quietly reaches 100

[Somerandomtechyboi]

1 hour ago, Feconi said:

Lol I care  If three-section water cooling can keep the fervor of the i9-12900K in games around 65-75 degrees, then in the render the temperature quietly reaches 100

What the hell is 3 section water cooling?

 

and intel just runs hot on ambient cooling even if you got a beefy car radiator waterloop and not dinky pc rad waterloop, something to do with heat not being able to be removed fast enough from the cpu die

 

its fine to run upto tjmax, though at stock id be more inclined to assume 115c max rather than 100c especially on cpus/gpus with low vcore, at that point youd be overheating everything else aswell cause the cpu heat might spread to the vrms or the rams

[Feconi]

5 minutes ago, Somerandomtechyboi said:

What the hell is 3 section water cooling?

 

and intel just runs hot on ambient cooling even if you got a beefy car radiator waterloop and not dinky pc rad waterloop, something to do with heat not being able to be removed fast enough from the cpu die

 

its fine to run upto tjmax, though at stock id be more inclined to assume 115c max rather than 100c especially on cpus/gpus with low vcore, at that point youd be overheating everything else aswell cause the cpu heat might spread to the vrms or the rams

in my country, three-section cooling is called cooling with a radiator on which 3 fans are installed

 

[Somerandomtechyboi]

33 minutes ago, Feconi said:

in my country, three-section cooling is called cooling with a radiator on which 3 fans are installed

 

Basically a dinky 360mm rad?

Ofc thatll struggle with a 12900k especially the garbage <30mm thick ones but itll do for stock operation or undervolt

[SorryBella]

8 hours ago, filpo said:

in my head it seemed to work 

 

Then again my head thinks that linus can go a whole vid without dropping anything

Yeah your head needs schizoid pill.

 

The main weaklink in an aircooling heatsink especially those with heatpipes are a phenomenon called dryout. In this case, the heat is so immensely higher than the dissipation capacity of the heatsink, that the capillary liquid inside of a heatpipe wont condense which destroys its performance but is very much reversible unless in an absolute utter catastrophic nightmare where the heatpipe blows out. But trust me, that is about as likely as you getting murdered by a school of shark.