What are good temps for the CPU & GPU while gaming or just under high loads?

[Azurael]

So I have been tooking at temps on my laptop while gaming for as long as I have owned it and tried many different things to keep them reasonable, but they will still spike at times. I know it is a laptop and cooling with a laptop is not easy. As such I have gotten used to seeing temps from low 70s (less intense games) up to high 80s (more intense games) on the CPU. Every now and then I have seen it spike for a couple minutes into the 90s, but that does not happen often. The GPU generally keeps temps around 68-76 in most cases I have thankfully not seen those spike much at all.

 

Now that I am getting a desktop I should be able to do more to keep the temps reasonable. I have gotten so complacent with the temps I get on my laptop (I really can't change much more than I already have here) I want to check with you all what are temperature ranges I should be shooting for on my CPU and GPU.

 

Case: Raimax x603 Mesh front (3 120mm fans in front)

CPU: 13700KF

CPU Cooler: DEEPCOOL LS720

GPU: 4070 ASUS Duel

 

If you need to see the other specs for a better idea of what I am working with let me know.

[Hinjima]

18 minutes ago, Azurael said:

So I have been tooking at temps on my laptop while gaming for as long as I have owned it and tried many different things to keep them reasonable, but they will still spike at times. I know it is a laptop and cooling with a laptop is not easy. As such I have gotten used to seeing temps from low 70s (less intense games) up to high 80s (more intense games) on the CPU. Every now and then I have seen it spike for a couple minutes into the 90s, but that does not happen often. The GPU generally keeps temps around 68-76 in most cases I have thankfully not seen those spike much at all.

 

Now that I am getting a desktop I should be able to do more to keep the temps reasonable. I have gotten so complacent with the temps I get on my laptop (I really can't change much more than I already have here) I want to check with you all what are temperature ranges I should be shooting for on my CPU and GPU.

 

Case: Raimax x603 Mesh front (3 120mm fans in front)

CPU: 13700KF

CPU Cooler: DEEPCOOL LS720

GPU: 4070 ASUS Duel

 

If you need to see the other specs for a better idea of what I am working with let me know.

The laptop temps were resonable, high but resonable for a laptop.  I'd say between 60-80c on the CPU ( 13700k is a hot chip ) and around 50-70c on the GPU depending on the game, some unoptimized games really hit the GPU hard.  My 4080 is usually at 55-60c at 30% fanspeed in any game like Metro Exodus and Dying Light 2.  Construction simulator 2022 however.. 77c at 65% fan speed!  While it looks like absolute garbage. Weird game.

[thrasher_565]

this is a good vidio about cpu. but even thow cpus only have a 3 year warranty they dont seem to just die. unless its the new stuff... but as the nm gets smaller more problems happen so more failers. but for most people it last long enough they wil most likey upgrade. theres vary few people that will use a pc till it dies thow normaly only one part diys...

 

but spending x amount of moeny to keep x part alive also is not vary cost effective ether as cpu and gpu come out like monthly so prices drop pritty fast.

 

however something thats like a collectors item and want to make it last also long as posable like water cooling a voodoo 5500 for example.

 

laptops on the other hand are well you get what you get. i dont see the point in them when all they do is thermlthrottl and they run hot and will diy faster.

 

back in the day 75c is about the norm. thow newer cpus are more in the 85c range and well in the video exspnsy why that is. there going more over the red zone.

 

 

Edited May 14, 2023 by thrasher_565

[Mark Kaine]

between 50 and 60c, everything higher i consider as hot!

 

Also weird question, maybe you should make that dependent if you experience issues or not?

[cmndr]

Lower is better but in practice it doesn't really matter, assuming you're not removing power limits or overvolting. 
If there's an issues, the chip will automatically down throttle.

Chips can last for years or even decades at high operating temps. 

I know someone who had the cooler on their desktop come undone. It ran for years effectively without a heatsink. Some crashing here and there. The unit's PSU died before the CPU. 

 

1 hour ago, Mark Kaine said:

between 50 and 60c, everything higher i consider as hot!

 

Also weird question, maybe you should make that dependent if you experience issues or not?

In 2003 maybe. Materials engineering has come A LONG way since then. It's been 20 years. 

[Mark Kaine]

32 minutes ago, cmndr said:

In 2003 maybe. Materials engineering has come A LONG way since then. It's been 20 years. 

No, nothing to do with that, at around that temp things are starting to boost lower,  fans getting more audible etc... ideal temp would be more like 30c tbh, but i consider 50-60 still acceptable  if your pc is running hotter (on average)  then its a hotbox and im sorry for your experience  = )

 

 

Edit: also i disagree,  yes a pc can last a decade,  but thats not the norm, especially "gaming pcs" aren't built to last, thermal paste, thermal pads, vrms, even the chips themselves,  come on people upgrade their stuff all the time, especially gpus often start to get slower and show issues after a couple of years... yes with good maintenance etc it can last , but its certainly defacto not the norm and heat isn't good for electronic parts, never will be

[freeagent]

Idling at 40-50c is terrible, and a 60c idle is unacceptable

 

If its not doing anything, you should be in the 20s-30s.

[Mark Kaine]

4 minutes ago, freeagent said:

Idling at 40-50c is terrible, and a 60c idle is unacceptable

 

If its not doing anything, you should be in the 20s-30s.

when i turn my pc on its 30c... for like 5 seconds... it does actually idle around 40c though,  so yeah not too bad and the fans are nearly inaudible at that point. 

[freeagent]

44 minutes ago, Mark Kaine said:

when i turn my pc on its 30c... for like 5 seconds... it does actually idle around 40c though,  so yeah not too bad and the fans are nearly inaudible at that point. 

I guess it isn't terrible.. a little warmer than what I am used to seeing.. this is just internetting..

 

[freeagent]

Well.. I guess that little tidbit doesn't paint a good picture..

 

[Blue4130]

2 hours ago, freeagent said:

Idling at 40-50c is terrible, and a 60c idle is unacceptable

 

If its not doing anything, you should be in the 20s-30s.

It completely depends on the ambient temperature. There is no hard and fast rule. It should be stated as delta over ambient.

[Blue4130]

2 hours ago, Mark Kaine said:

No, nothing to do with that, at around that temp things are starting to boost lower,  fans getting more audible etc... ideal temp would be more like 30c tbh, but i consider 50-60 still acceptable  if your pc is running hotter (on average)  then its a hotbox and im sorry for your experience  = )

 

 

Edit: also i disagree,  yes a pc can last a decade,  but thats not the norm, especially "gaming pcs" aren't built to last, thermal paste, thermal pads, vrms, even the chips themselves,  come on people upgrade their stuff all the time, especially gpus often start to get slower and show issues after a couple of years... yes with good maintenance etc it can last , but its certainly defacto not the norm and heat isn't good for electronic parts, never will be

The main reason for things slowing down is programs get bloated not hardware gets less capable. This has been proven many times. (assuming you clean the dust out)

[cmndr]

7 hours ago, Mark Kaine said:

No, nothing to do with that, at around that temp things are starting to boost lower,  fans getting more audible etc... ideal temp would be more like 30c tbh, but i consider 50-60 still acceptable  if your pc is running hotter (on average)  then its a hotbox and im sorry for your experience  = )

 

 

Edit: also i disagree,  yes a pc can last a decade,  but thats not the norm, especially "gaming pcs" aren't built to last, thermal paste, thermal pads, vrms, even the chips themselves,  come on people upgrade their stuff all the time, especially gpus often start to get slower and show issues after a couple of years... yes with good maintenance etc it can last , but its certainly defacto not the norm and heat isn't good for electronic parts, never will be

In this instance, you're out of date in your thinking. 
What you're saying was VERY mainline back in 2000. 
While cooler IS better, the benefit generally isn't meaningful anymore. You'd get a good cooler for less noise moreso than better performance/reliability.


For the last 5ish years every latptop has been designed with the assumption of the CPU regularly hitting 100C. What do you know that Apple and Dell don't?
Intel's box cooler is a joke... but they're not worried about overheating or significant amounts of degradation. What do you know that Intel doesn't?

 

In the case of servers, they're often running at heavy load and need MUCH higher levels of reliability than a normal person would. 

 

For what it's worth I used to work in an engineering/ops role at an F100 where the focus was on maintaining 6-9s of availability - very much the "1 minute of downtime is $1 million in cost" setting. CPUs pretty much never died. It was [software, personnel, PSUs, motherboards, RAM]. I don't think I ever saw a ticket mentioning CPUs and I skimmed through THOSUANDS of tickets since I was the numbers guy. 

 

 

----

 

https://www.intel.com/content/www/us/en/support/articles/000005597/processors.html#:~:text=The maximum junction temperature limit,details for specific processor models.

 

Quote

It's unlikely that a processor would get damaged from overheating, due to the operational safeguards in place. Processors have two modes of thermal protection, throttling and automatic shutdown. When a core exceeds the set throttle temperature, it will reduce power to maintain a safe temperature level. The throttle temperature can vary by processor and BIOS settings. If the processor is unable to maintain a safe operating temperature through throttling actions, it will automatically shut down to prevent permanent damage.

 

Not necessarily. Many Intel® processors make use of Intel® Turbo Boost Technology, which allows them to operate at very high frequency for a short amount of time. When the processor is operating at or near its maximum frequency it's possible for the temperature to climb very rapidly and quickly reach its maximum temperature. In sustained workloads, it's possible the processor will operate at or near its maximum temperature limit. Being at maximum temperature while running a workload isn't necessarily cause for concern. Intel processors constantly monitor their temperature and can very rapidly adjust their frequency and power consumption to prevent overheating and damage.

 

Could you provide a source from a CPU manufacturer, from the last few years, backing your claims?

[freeagent]

4 hours ago, Blue4130 said:

It completely depends on the ambient temperature.

For me it was 21c.

[Azurael]

7 minutes ago, freeagent said:

For me it was 21c.

Your ambient temp was 21c or the CPU idle temp was 21c?

[Mark Kaine]

3 hours ago, cmndr said:

In this instance, you're out of date in your thinking. 

Maybe i have different priorities? 

 

3 hours ago, cmndr said:

PSUs, motherboards, RAM].

*and GPUs

 

3 hours ago, cmndr said:

Could you provide a source from a CPU manufacturer, from the last few years, backing your claims?

i wasn't talking about CPUs specifically, i was talking overall PC temps and of course CPU plays a role in that... 

 

SSDs, mobos, psus, gpus, etc are all affected by heat, plus lower temps means lower noise obviously.  I mean seriously especially GPUs seem to be affected by that (and if its just the cooling solution,  but chips also dont seem to last forever) 

 

 

3 hours ago, cmndr said:

Could you provide a source from a CPU manufacturer,

Why would a CPU manufacturer say "yes, our chips run way too hot..." lol, and see above i wasn't talking about CPUs specifically anyway (not intentionally at least )

 

 

So Tldr: hot CPU/GPU = obviously bad for the system not at least because of downclocking shenanigans which everyone should want to avoid if possible ... longevity may be affected but thats not my main concern and i didn't mean to imply that.

[Vishera]

When it comes to CPUs and GPUs up to 85C is my recommendation for safe temperatures.

Even though the newer AMD CPUs max operating temperature is 95C, it doesn't mean that it's a good idea.

[freeagent]

2 hours ago, Azurael said:

Your ambient temp was 21c or the CPU idle temp was 21c?

My ambient was 21c. Did you not see my screenshot above.

[Mihle]

20 hours ago, Mark Kaine said:

 

 

SSDs, mobos, psus, gpus, etc are all affected by heat, plus lower temps means lower noise obviously.  I mean seriously especially GPUs seem to be affected by that (and if its just the cooling solution,  but chips also dont seem to last forever) 

That's not actually how it works, if it's higher temperature, it would mean your fan spins less fast, and that means less noise.

 

Let's say your CPU is at 70c stable at one fan RPM, if you let it become 80c instead because you run the fan a little less, that fan produce less noise.

20 hours ago, Mark Kaine said:

Why would a CPU manufacturer say "yes, our chips run way too hot..." lol, and see above i wasn't talking about CPUs specifically anyway (not intentionally at least )

As long as a CPU or GPU survives the first while, it's rare for the chip itself to die, it's 99% either the mobo or in GPUs case, the board it's attached to(and other things on it.

[BetteBalterZen]

On 5/13/2023 at 8:30 PM, Azurael said:

So I have been tooking at temps on my laptop while gaming for as long as I have owned it and tried many different things to keep them reasonable, but they will still spike at times. I know it is a laptop and cooling with a laptop is not easy. As such I have gotten used to seeing temps from low 70s (less intense games) up to high 80s (more intense games) on the CPU. Every now and then I have seen it spike for a couple minutes into the 90s, but that does not happen often. The GPU generally keeps temps around 68-76 in most cases I have thankfully not seen those spike much at all.

 

Now that I am getting a desktop I should be able to do more to keep the temps reasonable. I have gotten so complacent with the temps I get on my laptop (I really can't change much more than I already have here) I want to check with you all what are temperature ranges I should be shooting for on my CPU and GPU.

 

Case: Raimax x603 Mesh front (3 120mm fans in front)

CPU: 13700KF

CPU Cooler: DEEPCOOL LS720

GPU: 4070 ASUS Duel

 

If you need to see the other specs for a better idea of what I am working with let me know.

Anything within TjMax is techincally "good" temperatures, because chips are made to run within TjMax temperatures. As an example, if you run your CPU at 4GHz and under load hitting 60C, you're actually leaving performance on the table. In theory, it's much better (in terms of performance alone) to run 5GHz with loads temperatures nearing 90C. 

I see the video above from Der8aur, with the Intel technician he is interviewing. The Intel guy explains what I say above. 

All of this said though, I also prefer to have very cool chips because with cool chips you have lower fan RPM and lower fan RPM means less noise and dust, while also not pushing a CPU to the max, you don't use as much power overall. 

[thrasher_565]

39 minutes ago, BetteBalterZen said:

Anything within TjMax is techincally "good" temperatures, because chips are made to run within TjMax temperatures. As an example, if you run your CPU at 4GHz and under load hitting 60C, you're actually leaving performance on the table. In theory, it's much better (in terms of performance alone) to run 5GHz with loads temperatures nearing 90C. 

I see the video above from Der8aur, with the Intel technician he is interviewing. The Intel guy explains what I say above. 

All of this said though, I also prefer to have very cool chips because with cool chips you have lower fan RPM and lower fan RPM means less noise and dust, while also not pushing a CPU to the max, you don't use as much power overall. 

well that's because in short boots its better to get the work done fast and then ramp down then to run a little cooler but longer i think is the argument. and they know that there cpus dont just die and most people oc past what they recommend so... why not them oc?

 

ya i agree that cpus just dont die. posably degrade but the test was small so plasable 

 

for gpus thow about 60c for 24/7 use thow and no idea about cpus probly about the same?

[BetteBalterZen]

3 hours ago, thrasher_565 said:

well that's because in short boots its better to get the work done fast and then ramp down then to run a little cooler but longer i think is the argument. and they know that there cpus dont just die and most people oc past what they recommend so... why not them oc?

 

ya i agree that cpus just dont die. posably degrade but the test was small so plasable 

 

for gpus thow about 60c for 24/7 use thow and no idea about cpus probly about the same?

I read what you said mate but not sure I understand the point you're getting at?

But in terms of 24/7 usage; simply look up TjMax and you will know what temperatures you should be in, for 24/7 usage (technically). If TjMax is 100C, if you stay at 85-90C all day long, 24/7, then you're good. 

[thrasher_565]

4 hours ago, BetteBalterZen said:

I read what you said mate but not sure I understand the point you're getting at?

But in terms of 24/7 usage; simply look up TjMax and you will know what temperatures you should be in, for 24/7 usage (technically). If TjMax is 100C, if you stay at 85-90C all day long, 24/7, then you're good. 

i no gpus were diying running 100% when mining so they run them around 60%?

[BetteBalterZen]

11 minutes ago, thrasher_565 said:

i no gpus were diying running 100% when mining so they run them around 60%?

I am not sure but I think this specifically was VRAM temperatures becoming an issue but I may be wrong. 

[thrasher_565]

Just now, BetteBalterZen said:

I am not sure but I think this specifically was VRAM temperatures becoming an issue but I may be wrong. 

oh i think that's what mining uses i think. so gpu core is fine 100% and vrm 60% then? unless cooled better?