Hello, I tried every possible solution (listed in bottom) but it just doesn't work (unless I disable PBO which I don't want I get 75 Celsius at 3.7GHz). I though it is cooler problem so I bought Be quiet! DARK ROCK SLIM but the results disappointed me. I still got 95 degree Celsius (it even hit 97) after 45 min stress test with maximum frequency of 4.3GHz and avarage about 4.1GHz. Is it that the cooler isn't enough for the CPU (which I read it should be fine) or could it be CPU/MB issue? The issues were same with new PC and ASUS support said that the temperatures are normal...
Specs:
I tried:
T0oKEN, what are you using to measure temperature (screenshot please)? In general re-pasting seldom helps and exposes possible drastic failures like processor stuck to cold plate. John.
HWmonitor + OCCT for benchmark, AMD adrenaline have same reports so I will only send you from these two
Thanks, T0oKEN, does Adrenalin ever report 95.8C? I have always suspected that AMD Ryzen Master (RM) knows something the rest of us do not. The maximum temperature (TjMax) is 95C. HWm does not tell us what Core temperature is but then I do not use it since it occasional reports absurd temperatures. If you ever use Windows, please post a screenshot of RM running Cinebench R24 Multicore. John.
@T0oKEN is this a new build or existing that temperature changed?
My son's 5600 was recently indicating 95°C via CoreTemp, whereas previously it ran around 75°C. Must have been a chipset or Windows update, cause after updating CoreTemp to latest version it was back to reporting normal temp.
Also, you mention using PBO, assuming for higher boost? Do you also have Curve Optimizer enabled? Our 5600 runs stable at -30 CO and runs a lot cooler as a result.
I tried Curve Optimizer running stable at -20 but the temps are still hitting 95 degree Celsius but after longer period of time (from basically instantly to one minute). I can try running up linux and do benchmark here to see any difference.
Edit: forgot to mention that I got higher boost from 4.1GHz to 4.4GHz
2nd edit: haven't answered the first question so yes, it was the same so that why I bought better cooler because the original one in PC was I think stock AMD
Be quiet! DARK ROCK SLIM is probably an adequate cooler for stock frequencies but when overclocking better cooling will help.
My son's 5600 overclocked gets toasty and that's even with a 240mm AIO.
OK so i tried stress test on Linux mint and got 88 Celsius at 4.4GHz. Still weirdly high but it's better.
FunkZ, CoreTemp along with most other non-AMD utilities often report erroneous temperatures. I use AIDA64 which is accurate but a paid application so I do not usually recommend it. I do not know where/when HWmonitor gets its maximum temperatures but I have always suspected it. Oh well, John.
The max temperatures are the highest temperature it hits
Of course, but perhaps RM knows to ignore erroneous values. Does one sample of high temperature in an hour mean the processor is running at that temperature? You are free to use any utility you choose. Enjoy.
For troubleshooting you should reset cmos, forget about pbo/cpb for the moment.
The Voltages, Powers readouts are too high. Is windows power plan on balanced?
Here's mine using cpu-z stress (bios changes - spread spectrum disabled, cool & quiet disabled, xmp profile on, the rest untouched),
With cmos restarted, it does disable PBO and D.O.C.P which runs under 80C. I'm currently trying to understand where is problem with it enabled since the cooler should be enough to cool it. You lose on performance with it disabled.
And what are the Voltages, Power readouts with that? Post an image.
What motherboard model is used in that prebuilt pc, and are you using any 3rd party, or asus oc tweaking software?
I already posted the image. The motherboard is PRIME B550M-K. I tried both using only bios and then tried using on an original prebuild ASUS Armoury Crate and on upgraded PC AMD Adrenalin and AMD Ryzen Master. All results are same.
Oh, i thought that image was from with pbo, cpb enabled, not baseline without them in use for comparison.