[SOLVED] Tjmax vs thermal throttling

Apr 30, 2021
36
0
30
0
Ryzen master indicates that the thermal throttling of 5900x is 90C, while in many threads people say that the Tjmax of 5900x is also 90C.
I mean, thermal throttling is the threshold for reducing the power consumption of the CPU, while Tjmax is the maximum temperature that the CPU should be on.

Doesn't it make more sense to make the thermal throttling lower slightly lower than Tjmax, say 85C, so it will NEVER reach 90C?
 
Ryzen master indicates that the thermal throttling of 5900x is 90C, while in many threads people say that the Tjmax of 5900x is also 90C.
I mean, thermal throttling is the threshold for reducing the power consumption of the CPU, while Tjmax is the maximum temperature that the CPU should be on.

Doesn't it make more sense to make the thermal throttling lower slightly lower than Tjmax, say 85C, so it will NEVER reach 90C?
It doesn't really matter because because in neither case it should be allowed to reach 90c.
Technically Tjmax is temperature that you can measure using a thermo-couple embedded in the center of CPU. Reason is that not only it will severely throttle to a point of shutting down affected cores but 10 or so degrees before that boost will be also lowered in case of automatic boost settings.
 
Ryzen master indicates that the thermal throttling of 5900x is 90C, while in many threads people say that the Tjmax of 5900x is also 90C.
I mean, thermal throttling is the threshold for reducing the power consumption of the CPU, while Tjmax is the maximum temperature that the CPU should be on.

Doesn't it make more sense to make the thermal throttling lower slightly lower than Tjmax, say 85C, so it will NEVER reach 90C?
It doesn't really matter because because in neither case it should be allowed to reach 90c.
Technically Tjmax is temperature that you can measure using a thermo-couple embedded in the center of CPU. Reason is that not only it will severely throttle to a point of shutting down affected cores but 10 or so degrees before that boost will be also lowered in case of automatic boost settings.
 
Apr 30, 2021
36
0
30
0
Thanks for your comment! So the Tjmax is already lower than the actual heat limit of the CPU, and reaching the Tjmax in some peaks doesn't harm the CPU, right?

I already lowered the thermal throttling to 88C in the BIOS, since I prefer to keep it cooler and add some margin from Tjmax. This processor is a beast, it boosts really high and increases its temperature in a second. I've already seen some 90C peaks.. it can reach 80C on load, but when it needs slightly more power, it reaches up to 90C (and I don't like it :) ). If it's not an issue, I will return it to 90C.
 
Last edited:
Thanks for your comment! So the Tjmax is already lower than the actual heat limit of the CPU, and reaching the Tjmax in some peaks doesn't harm the CPU, right?

I already lowered the thermal throttling to 88C in the BIOS, since I prefer to keep it cooler and add some margin from Tjmax. This processor is a beast, it boosts really high and increases its temperature in a second. I've already seen some 90C peaks.. it can reach 80C on load, but when it needs slightly more power, it reaches up to 90C (and I don't like it :) ). If it's not an issue, I will return it to 90C.
What you are seeing is highest temp at hottest core not necessarily whole CPU and even if one core throttles, typically only one core may not have full boost or even throttle, rest of cores should pick up the load and so you may loose only a fraction of performance. In any case cooler it is, better performance.
 
Apr 30, 2021
36
0
30
0
I'm actually monitoring the CPU Die average with HWiNFO64. It can peak up to 90C. Ryzen Master monitors the highest sensor, and it also can reach 90C. But the CPU Die average more worries me...
 

ASK THE COMMUNITY