• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

i5-10300H Undervolting, Question

Joined
May 3, 2021
Messages
7 (0.01/day)
Is your CPU running at full speed or is it throttling? Are you using ThrottleStop 9.3? Limit Reasons data was not displayed for the 10th Gen CPUs in previous ThrottleStop versions.

If you have any questions, post some screenshots of ThrottleStop while the CPU is loaded with Limit Reasons open.

No, it was running on a 43, 41,41,41 (turbo ratios). Now it's on a 45, 45, 45 and 40, just as above. Quite an interesting setup.

And yes, I'm using TS 9.3, couldn't have done all this without it (Intel's utility is a not the kind of tool I'm inclined to use).

What I meant was that all is fine, fortunately for me there are no limit reasons showing up and no thermal/power/voltage based throttling. Hence, what I said, and that I expect no LRs to show up, given how stable this processor has been.

The question was more geared towards HWInfo64. I can't for the life of me figure out why would it clear flags, and thus makes me a bit uncomfortable using it.
 

nikola87

New Member
Joined
Nov 26, 2020
Messages
18 (0.01/day)
@nikola87

This ^^^ (ABOVE) actually produced interesting results.

PL1 I set it to 60 as originally suggested by unclewebb, actually before I also kept it at 55, this is because during the test with Cinebench after a few minutes the processor becomes more hungry for energy and 45W is not enough, so it is already satisfied at 50, but it is a cpu that can work well even at 55W

I have left the PL2 unchanged, a lower value can make the warning light pl1/2 turn yellow and red again for low power speech.

Not even me in HWiNFO64 are shown PL1 and PL2 but I had already noticed for quite a while because without TS they reappear

So you can take it easy
 
Last edited:

unclewebb

ThrottleStop & RealTemp Author
Joined
Jun 1, 2008
Messages
8,134 (1.33/day)
all is fine,
I was worried that something was broken when you said that nothing was showing up in Limit Reasons. No throttling is a good thing.

HWiNFO64 manages the throttling flags internally. This process involves clearing the throttling data out of the CPU as it happens. ThrottleStop only clears this data if a user presses the Clear button in Limit Reasons.

I can understand why HWiNFO64 does this automatically. That gets rid of the yellow warning boxes that ThrottleStop shows. Users always ask, what do those yellow boxes mean. Just after booting up or resuming, yellow boxes often times do not mean anything. Some users might think that clearing this data is a good thing.
 
Joined
May 3, 2021
Messages
7 (0.01/day)
I was worried that something was broken when you said that nothing was showing up in Limit Reasons. No throttling is a good thing.

HWiNFO64 manages the throttling flags internally. This process involves clearing the throttling data out of the CPU as it happens. ThrottleStop only clears this data if a user presses the Clear button in Limit Reasons.

I can understand why HWiNFO64 does this automatically. That gets rid of the yellow warning boxes that ThrottleStop shows. Users always ask, what do those yellow boxes mean. Just after booting up or resuming, yellow boxes often times do not mean anything. Some users might think that clearing this data is a good thing.

Indeed (no throttling is good).

Regarding HWINFO64: that justification is plausible, but I prefer your approach: inform the user CLEARLY, clear remaining doubts in the right forums, and make clearing "warnings" an option. That way it's the user who's in control the whole time.

Ah, btw, I had wondered if TechPowerUp GPU-Z did the same HWINFO64 does, but the information remains visible (in MAX STATE) until the user resets the whole monitoring.


Thank you for helping out, unclewebb.

@nikola87:
All good.

Managing desktops requires dedication (proper selection of components, proper cooling, ordering the cables inside the case) and the usual monitoring of all temps and fans.

But with powerful laptops it's constant worry-time, to be honest. I have been lucky with my HP Omen, they did a good job (surprisingly) with the cooling system, but still temps go a bit high without proper intervention (good old undervolting of cpu/gpu).

But all good.
 

nikola87

New Member
Joined
Nov 26, 2020
Messages
18 (0.01/day)
I would like to make a report, lately it seems that the beta version 9.3.1 the UV on my cpu no longer had any effect it was as if it were not there, 100 degrees during the tests practically immediately, before drawing other conclusions I put back the stable 9.3 and everything seems to be back as before and now it doesn't go beyond 90 degrees under stress, I was starting to worry and repasting as a last resort, although I had a raised base under the laptop it felt very strange to me
 
Last edited:

unclewebb

ThrottleStop & RealTemp Author
Joined
Jun 1, 2008
Messages
8,134 (1.33/day)
There should be no difference between TS 9.3 and TS 9.3.1. They are equally stable and both versions can apply the CPU voltages correctly. If you are seeing a difference in temperatures during the exact same stress test then there is something else going on that is causing that. Post some screenshots of each version and let me know what stress test you are running that shows a repeatable temperature difference.

Have a look in the FIVR monitoring table and make sure your offset voltages are being applied to both the core and to the cache. A random Windows update can instantly disable CPU voltage control in the blink of an eye.

It is normal for laptop temperatures to get worse over time. Some thermal pastes degrade, room temperatures get hotter, etc.
 

nikola87

New Member
Joined
Nov 26, 2020
Messages
18 (0.01/day)
There should be no difference between TS 9.3 and TS 9.3.1. They are equally stable and both versions can apply the CPU voltages correctly. If you are seeing a difference in temperatures during the exact same stress test then there is something else going on that is causing that. Post some screenshots of each version and let me know what stress test you are running that shows a repeatable temperature difference.

Have a look in the FIVR monitoring table and make sure your offset voltages are being applied to both the core and to the cache. A random Windows update can instantly disable CPU voltage control in the blink of an eye.

It is normal for laptop temperatures to get worse over time. Some thermal pastes degrade, room temperatures get hotter, etc.

I redid the tests with TS bench with both versions and this time both were fine, temperatures ok, maybe then in those moments there must have been some conflict with a program or something like that, maybe the culprit is me who should be fiddling less :p
 

unclewebb

ThrottleStop & RealTemp Author
Joined
Jun 1, 2008
Messages
8,134 (1.33/day)
The TS Bench is fine for creating some load and for creating some heat. It is best to use a completely separate benchmark program when doing some serious testing. Something like Cinebench R20 should be fairly consistent as long as your CPU is not thermal or power limit throttling.

 
Top