Performance i7-10875h - too low? (benchmarking) [solved by reinstalling windows]
- Thread starter Vince92
- Start date
FerrariVie
Super Star
Great! Even now, you won't find any difference when changing the profiles in gaming center? You should get a considerable smaller score in cinebench when using office mode, comparing to gaming mode or turbo (benchmarking mode).
Your CPU temps are too high now, so I'm assuming you're using turbo mode?
Your CPU temps are too high now, so I'm assuming you're using turbo mode?
barlew
Godlike
Hi everyone, thank you for your help. Unfortunately I did not find out who the culprit was BUT the reinstallation has been successful:
From 2400 to 4072 points in Cinebench R20
It is also getting immensely hotter and louder
View attachment 20643
View attachment 20644
View attachment 20645
Much more like it mate. Glad you got it sorted.
It was Gaming Mode. I will test all three modes and log them to compare. But first I install all my programs back
ubuysa
The BSOD Doctor
It seems you're happy, but for the record that device is the Nvidia USB Type-C Port Policy Controller. It's the Nvidia VirtualLink port.
SpyderTracks
We love you Ukraine
Oh, I'm so pleased that's performing better! I had an inkling it was a configuration issue.
FerrariVie
Super Star
Did you reset the values in Hwinfo between each mode? If yes, then reaching 95° while on office mode doesn't seem right to me, but I don't have this chassis or CPU, so not sure if that's kind of expected.
@barlew , you at least have the same CPU, right? Even though I know your chassis is 17", but do you have similar temps at stock, even when using office mode?
barlew
Godlike
I do yes. Ill give it a go later and report back.
Those temps in office mode don't surprise me as the laptop massively ramps down fan usage.
barlew
Godlike
For your awareness on this CPU the max you can really have a stable undervolt is around 0.70mV and even that is pushing it.
barlew
Godlike
By the way my gaming performance also increased.
Before the reinstallation of windows I got 100 to 115 fps in COD BO Cold War.
Now I get 120-140 fps and this also confirms that the throttling was not only applied to the benchmarking.
Right now I capped at -106,4 mV undervolting of the CPU. -108 an lower is not stable anymore.
@barlew I saw -70 mV in the BIOS as this is the maximum you can go.
Before the reinstallation of windows I got 100 to 115 fps in COD BO Cold War.
Now I get 120-140 fps and this also confirms that the throttling was not only applied to the benchmarking.
Right now I capped at -106,4 mV undervolting of the CPU. -108 an lower is not stable anymore.
@barlew I saw -70 mV in the BIOS as this is the maximum you can go.
FerrariVie
Super Star
You could get even more, because if your CPU is reaching 95, then it is thermal throttling. By how much we don't know, but it could definitely be better. However, since you've already undervolted, then the only solution left is a repaste with a better thermal compound (like TG Kryonaut or Noctua NT-H2).
But if you're happy with current performance and noise, then don't worry about it
barlew
Godlike
You could get even more, because if your CPU is reaching 95, then it is thermal throttling. By how much we don't know, but it could definitely be better. However, since you've already undervolted, then the only solution left is a repaste with a better thermal compound (like TG Kryonaut or Noctua NT-H2).
But if you're happy with current performance and noise, then don't worry about it
I am going to disagree here. Those 95's are maximum values which are the spike values created when the benchmark starts from idle. His average temps are absolutely fine. Repasting may get the average temps down but I reckon those spikes will maintain.
FerrariVie
Super Star
I get your point, but what I'm saying is that as soon as you hit 95, there is thermal throttle going on for sure. But if it is really only spikes of a few seconds, then the performance hit won't be huge.
Edit: You can't know the duration of the 95° just by looking at averages, you need to log the cinebench session (or gaming, even better) and see how long it stays there in 95° (or keep looking at the temperatures while it's running, which works for benchmarks but not really for real-world usages, like gaming).
barlew
Godlike
I see your reasoning and if the average temps were considerably higher I would agree with you. It would indicate the spikes were lasting a considerable amount of time which would induce severe thermal throttling as you said.
The average temps shown indicate that these are just initial spikes occurring before the fans kick in. I would be surprised if they last for longer than a second.