[SOLVED] 97c on Ryzen 7 5700G

[elseq3]

Hi, I had recently been playing Cyberpunk 2077 and noticed that my FPS had dropped, then I opened Afterburner and noticed that my processor (Ryzen 7 5700G) was reaching temperatures up to 92 degrees celsius, so I preferred not to play Cyberpunk for now, as no other game takes my temperatures to this extreme.

However, I tested the latest version of Cinebench and noticed when I opened HWMonitor that my temperature reached 97 degrees, I don't know how long it was like that but I know I had the test open for a few minutes, obviously after realizing my temperatures were so high, I closed the test.

Do you think I have caused damage to the processor? Just now everything is working as usual, but I'm afraid that my processor will stop working or something.

 

[hotaru.hino]

The only thing remaining at extreme temperatures does is wear out the processor faster. But there's no hard equation or anything that says "if it's been like this for X days, you lose Y days" or whatever. Plus it'd have to be like this for a while to really affect its life expectancy.

In any case, if the computer is reasonably clean of dust and the cooler seems securely mounted to the CPU, check the thermal paste. It might've pumped out enough to stop doing its job effectively.

 

[elseq3]

The only thing remaining at extreme temperatures does is wear out the processor faster. But there's no hard equation or anything that says "if it's been like this for X days, you lose Y days" or whatever. Plus it'd have to be like this for a while to really affect its life expectancy.

In any case, if the computer is reasonably clean of dust and the cooler seems securely mounted to the CPU, check the thermal paste. It might've pumped out enough to stop doing its job effectively.

Thanks for your answer.
So there is no problem if it still exceeded Tjmax (95 °C) during those minutes running Cinebench?

I will do a preventive cleaning tomorrow and apply new thermal paste to see if the temperatures improve, I'm currently using Wraith Stealth cooler but I'm planning to buy the DeepCooler AK620 Zero Dark.

 

[hotaru.hino]

Thanks for your answer.
So there is no problem if it still exceeded Tjmax (95 °C) during those minutes running Cinebench?

I wouldn't say there isn't a problem, but I wouldn't worry about it. Especially if this only happened for a short period of time and you didn't continue pushing it

I will do a preventive cleaning tomorrow and apply new thermal paste to see if the temperatures improve, I'm currently using Wraith Stealth cooler but I'm planning to buy the DeepCooler AK620 Zero Dark.

That should definitely help with cooling.

 

[elseq3]

I wouldn't say there isn't a problem, but I wouldn't worry about it. Especially if this only happened for a short period of time and you didn't continue pushing it


That should definitely help with cooling.

Yes, I'm not planning on running Cinebench or Cyberpunk until I fix my cooling issues, I was just worried if I had caused any damage by reaching those temperatures in that period of time.

Thank you very much for your help, I really appreciate it, I'm calmer now.

 

[TerryLaze]

Look at vcore numbers, temperature is only a symptom, high temp comes from high power usage which goes together with high voltage.
It's the voltage that degrades a CPU and not the heat.
"Fixing" increasing the cooling could just make your mobo use even more extreme settings if you have PBO or something similar enabled.

If your system is running at normal specs and it's really the cooling that is the problem then that will also be shown by vcore because it will be pretty low.

 

[elseq3]

Look at vcore numbers, temperature is only a symptom, high temp comes from high power usage which goes together with high voltage.
It's the voltage that degrades a CPU and not the heat.
"Fixing" increasing the cooling could just make your mobo use even more extreme settings if you have PBO or something similar enabled.

If your system is running at normal specs and it's really the cooling that is the problem then that will also be shown by vcore because it will be pretty low.

Thank you so much for your answer, it's nice to see things from several points of view to fix my cooling problem.

These are currently my VCORE numbers.

Do they seem okay?

 

[hotaru.hino]

Thank you so much for your answer, it's nice to see things from several points of view to fix my cooling problem.

These are currently my VCORE numbers.

Do they seem okay?

If you left the voltage settings on stock, they're fine. While I'm not sure if this applies to APUs (of which yours is considered one) and with nothing else to go off of, AMD has said that Zen 3 CPUs are designed to handle up to 1.5V. See https://hardforum.com/threads/1-5v-...stilt-1-55v-is-the-cap-for-zen-3-amd.2015950/

 

[elseq3]

If you left the voltage settings on stock, they're fine. While I'm not sure if this applies to APUs (of which yours is considered one) and with nothing else to go off of, AMD has said that Zen 3 CPUs are designed to handle up to 1.5V. See https://hardforum.com/threads/1-5v-...stilt-1-55v-is-the-cap-for-zen-3-amd.2015950/

Sorry if I'm making a dumb comment, but just now I only make use of the processor, I disabled the integrated graphics as I recently installed an RTX 4060 to my PC.

 

[TechyInAZ]

As previously mentioned there shouldn't be anything wrong with the CPU going just 2 degrees higher. Its virtually impossible for a microchip processor to hold an exact perfectly linear temperature limit. Simply put, a CPU can't clock up or down fast enough to compensate. This is why GPUs can often consume double their TDP rating for a extremely brief period of time.

We don't know how AMD validates its CPUs, but its safe to say that AMD's maximum temperature limit will be below the silicon's actual thermal rating. I wouldn't be surprised if the CPU can actually withstand 100-110C for a lengthy period of time before death.

The CPU will also reduce clocks and voltage the higher your temperature is. Voltage + amps + heat is what kills CPUs.

 

[elseq3]

As previously mentioned there shouldn't be anything wrong with the CPU going just 2 degrees higher.

Thanks for answering. When I opened this thread I was scared because I thought that reaching those temperatures could cause damage to the processor, although as I mentioned, at most it was at 80 °C most of the time (since the highest load comes in the city, where the processor reached the temperature of 90 °C, but I was playing mostly some missions in more enclosed places, so I don't think it was always under load) while playing Cyberpunk for about five-six hours, and with Cinebench it won't have been more than 10 minutes (which is what set off my alarms because the temperature exceeded the Tjmax, as I mentioned), but now I understand that it's really not that accurate and just because that happens for a short period of time doesn't mean the processor is going to be damaged; right now my temperatures while gaming (and since I got my PC) are between 60 °C and 70 °C, and I'm waiting for the new fan cooler to improve them a bit more.

I wouldn't be surprised if the CPU can actually withstand 100-110C for a lengthy period of time before death.

Shouldn't the computer shut down before reaching those temperatures, or are you talking about a hypothetical case?

Voltage + amps + heat is what kills CPUs.

I also understand that, what also worried me is that when I proceeded to close Cinebench, I did not notice slowness or instability in the system although the temperatures have risen a lot, I guess what you mention does not mean that the performance will drop considerably but it does mean that the processor will be lowering its power and this I assume because while playing Cyberpunk the FPS had dropped.

 

[TechyInAZ]

I was talking about a hypothetical case. Should the CPU run at over 100C it should be able to do so without dying for at least some lengthy period of time. At least you would expect. Generally when it comes to product stability, you always want some wiggle room as an insurance policy. You don't want your products riding on the exact edge of stability or longevity.

Its worth mentioning that these chips are designed to operate at their TJMax or lower. If a chip is rated at 95C it means that it is designed to run at that temperature for at least several years, or until the warranty expires.

CPUs normally can operate well beyond their warranty periods as well.

For example, AMD's latest Ryzen 7000 and Ryzen 8000 CPUs are designed to hit their TJMax or 95C even under the most demanding cooling on the market, so that they can eek out as much performance as possible.

If it makes you feel any better i have a Ryzen 7 5800X3D with a beefy dual-tower 120mm heatsink, and it regularly goes up to 85-90C (which is its TJMax) all the time under extremely demanding workloads. I don't care that it does, in-fact i build my fan curves around those high temps so the fans only start spinning fast when it hits over 80-85C.

 

[TechyInAZ]

The only potential issues with high temps that are actually legitimate issues (potentially) is that you'll loose some boost clock which will reduce performance. But that's down to personal preference and how much performance you are willing to loose. Boost clocks are not guaranteed anyways so the CPU is still operating within spec.

Overclocking is also another problem. If you want to overclock at all, running the CPU at TJMax is really bad. You won't have a lot of overclocking headroom and you'll have no room to increase voltage.

But for stock operation, its totally fine.

 

[elseq3]

I was talking about a hypothetical case. Should the CPU run at over 100C it should be able to do so without dying for at least some lengthy period of time. At least you would expect. Generally when it comes to product stability, you always want some wiggle room as an insurance policy. You don't want your products riding on the exact edge of stability or longevity.

Its worth mentioning that these chips are designed to operate at their TJMax or lower. If a chip is rated at 95C it means that it is designed to run at that temperature for at least several years, or until the warranty expires.

CPUs normally can operate well beyond their warranty periods as well.

For example, AMD's latest Ryzen 7000 and Ryzen 8000 CPUs are designed to hit their TJMax or 95C even under the most demanding cooling on the market, so that they can eek out as much performance as possible.

If it makes you feel any better i have a Ryzen 7 5800X3D with a beefy dual-tower 120mm heatsink, and it regularly goes up to 85-90C (which is its TJMax) all the time under extremely demanding workloads. I don't care that it does, in-fact i build my fan curves around those high temps so the fans only start spinning fast when it hits over 80-85C.

Thank you very much for taking the time to explain that to me. How long have you had your processor running at those temperatures?

Overclocking is also another problem. If you want to overclock at all, running the CPU at TJMax is really bad. You won't have a lot of overclocking headroom and you'll have no room to increase voltage.

But for stock operation, its totally fine.

Well, I'm not planning to do any overclocking, I only had activated OC Genie on my CPU and it's currently running perfectly at good temps.

 

[TerryLaze]

I was talking about a hypothetical case. Should the CPU run at over 100C it should be able to do so without dying for at least some lengthy period of time. At least you would expect.

CPUs have baked in temp maximums at which point they will just shut down immediately, at 95° they just slow down to recover some temp headroom.
At least intel has that and it's documented, AMD doesn't have public data sheets but it should also have the same kind of protection.

Back in the days amd cpus would just fry on the spot.

View: https://www.youtube.com/watch?v=YYQSHXNFvUk&ab_channel=MikesTek

 

[elseq3]

CPUs have baked in temp maximums at which point they will just shut down immediately, at 95° they just slow down to recover some temp headroom.
At least intel has that and it's documented, AMD doesn't have public data sheets but it should also have the same kind of protection.

Back in the days amd cpus would just fry on the spot.

View: https://www.youtube.com/watch?v=YYQSHXNFvUk&ab_channel=MikesTek

Wow, I know it's an old video but if it takes that temperature for a processor to melt down, I don't think absolutely nothing would have happened by going to 97 °C for a short period of time.

 

[rcald2000]

Upgrade your cooler:

----------------------
RGB variant
Thermalright BA120 ARGB CPU Air Cooler, 6 Heat Pipes, TL-C12CG-S PWM Quiet Fan CPU Cooler with S-FDB Bearing, for AMD AM4/AM5 Intel LGA1700/1150/1151/1200
$22.90

https://www.amazon.com/dp/B09YY72PYN/

----------------------
Non-RGB variant
Thermalright BA120 CPU Air Cooler, 6 Heat Pipes, TL-C12C PWM Quiet Fan CPU Cooler with S-FDB Bearing, for AMD AM4 AM5/Intel 1700/1150/1151/1200, PC Cooler
$21.49

Amazon.com

----------------------

 

[elseq3]

Upgrade your cooler:

----------------------
RGB variant
Thermalright BA120 ARGB CPU Air Cooler, 6 Heat Pipes, TL-C12CG-S PWM Quiet Fan CPU Cooler with S-FDB Bearing, for AMD AM4/AM5 Intel LGA1700/1150/1151/1200
$22.90

https://www.amazon.com/dp/B09YY72PYN/

----------------------
Non-RGB variant
Thermalright BA120 CPU Air Cooler, 6 Heat Pipes, TL-C12C PWM Quiet Fan CPU Cooler with S-FDB Bearing, for AMD AM4 AM5/Intel 1700/1150/1151/1200, PC Cooler
$21.49

Amazon.com

----------------------

Thanks! I already ordered a AK620 ZERO DARK, it should arrive in two weeks.