Confused by Package/Ring Throttling and Power Limit readings

Cwell81

New Member
Hi Everyone,

I am relatively new to HWINFO and have only started using it to monitor and tweak the performance of my new laptop, which is an HP Omen 15 2020 model with an i7-10750H CPU, 16 GB RAM, and a Nvidia RTX 2060 GPU. I've included fuller specs below using system information. Let me know if additional info would help. And apologies in advance for what might be a stupid question, but I've found little in the way of explanation despite multiple Google, Reddit, and HWINFO forum searches.

In order to test temperatures while playing games under load, I fired up a copy of RAGE 2 and used Riva Tuner to display frame rates and CPU/GPU temps in game. I am limiting the boost on my CPU using Windows 10 power options (setting Processor Performance Boost Mode to "Efficient Aggressive") and setting the fan/performance profile to "Balanced" (which mainly determines how and when the fans ramp up). My goal is to get decent gameplay at high settings with CPU temps under 95 degrees. I figure if I can keep it under 90, I'll be in even better shape. I don't need MAX performance when gaming. At least not until I'm more comfortable with temp limits.

After playing for about a half hour, I noticed a couple of small temp spikes into the high 80s (Celsius), but generally observed temperatures of 85 degrees or less on the CPU and 72 degrees or less on the GPU. There were very few hiccups in gameplay and if frames were skipped, I didn't notice them. Frame rates were generally between 80 and 120 FPS.

Then, I decided to close out the game and check the readings from HWINFO. I was a little surprised to see a max CPU core temp of 94 degrees, but the average temp was 84, so I assumed it was a brief spike. Then I noticed two red readings in the sensor status window that I just can't explain. Both the Package/Ring Thermal Throttling and Package/Ring Power Limit Exceeded readings said "Yes." I checked all the temps on all of my cores and none of them were over 94, but two of the cores registered thermal throttling at some unknown point. I then tried to check how I might have exceeded the power limit on anything, but quickly realized I was way out of my depth.

I checked Intel's page for my CPU and noted that the Tjunction max is 100 degrees Celsius, which means I never should have throttled in the first place (I think). Performance was great all around, the chassis never got hot, and I was thoroughly satisfied with all the average numbers I saw, so what could be going on here? Does HWINFO register different throttle limits based on the power plans you have in place? Could the temps I'm seeing being wrong? Any help you can provide is much appreciated!

OS Name Microsoft Windows 10 Home
Version 10.0.18363 Build 18363
System Manufacturer HP
System Model OMEN Laptop 15-ek0xxx
System Type x64-based PC

Processor Intel(R) Core(TM) i7-10750H CPU @ 2.60GHz, 2592 Mhz, 6 Core(s), 12 Logical Processor(s) [NOTE, this boosts up over 4 GHz on a regular basis]
BIOS Version/Date AMI F.05, 7/27/2020
SMBIOS Version 3.2
Embedded Controller Version 17.24

Installed Physical Memory (RAM) 16.0 GB
Total Physical Memory 15.8 GB
Available Physical Memory 8.77 GB
Total Virtual Memory 18.6 GB
Available Virtual Memory 7.74 GB

Intel(R) UHD Graphics
Adapter RAM 1.00 GB (1,073,741,824 bytes)
Driver Version 26.20.100.7985
Resolution 1920 x 1080 x 300 hertz
Bits/Pixel 32

NVIDIA GeForce RTX 2060
Adapter Type GeForce RTX 2060, NVIDIA compatible
Adapter Description NVIDIA GeForce RTX 2060
Adapter RAM (1,048,576) bytes
Driver Version 27.21.14.5167
 
I am having a very similar issue with an i7-10750H in an Acer Predator. I wonder if this issue is specific to laptops? Have you found out any additional info?

The 2 for me that continue to warn are package/ring thermal and sometimes the power one. Both throttling.

One thing is when I set the turbo on which has a 100% fan all the time. This doesn't happen. The highest temps reached at 93 and I know that I should be able to hit 100 without any throttling.

Would be great to figure out if I can set the warning level as well as any throttling (if any is actually occurring) to a higher value.

Cheers,
T
 
I wouldn't be surprised this being tripped on notebooks as those devices are thermally constrained and often reach quite high temperatures. This is tolerated by the vendors to a certain degree as such systems are unable to provide sufficient cooling.
 
I was thinking the same. My particular model has the TURBO mode with a fan that is crazy loud; but seems to alleviate the issue. I was just surprised that this occurs so often. I mean even on boot ups. It actually happened again right now as I type this note on my desktop PC. My Laptop is just sitting here doing absolutely nothing. LOL...

Thanks for the response. I guess I will just turn these alarms off as it doesn't appear there is a way to lower their triggering frequency.

What would be great would be like a 1 minute threshold. I.E. only launch the alert is the condition exists for a particular length of time.

Anyway, thanks for the response.

Cheers,
Tamer
 
There are options in the Alarm settings that allow setting up such criteria.
 
I see Notification Distance and Samples to trigger. Should I trigger only after 20 samples? How often are samples taken?

Thanks,
T
 
That's determined by your polling frequency. If you have 2000 ms polling and set alarm to 30 samples, it will trigger it only when the value in exceeding the threshold for 1 minute (30 * 2 seconds).
 
Hey there. The thermal throttle being tripped just means that it used the algorithm that reduces the temperature somewhere in that group and is normal on occasion assuming that the power limits are set to ride the maximum line for a giving set of hardware and cooling solutions. Those Power Limits are the governor for the amount of power being used along with a current limiter. Those two limits are some of the primary methods to keep the system in balance and those limits being hit just means The CPU is doing some good work. :)

The TJ temp is settable, viewable and possibly set below 100C in a laptop by the OEM. Also T-junction temps are not the same as the CPU Package temp which is the number you are (likely) seeing at 94C and is a 250ms average of [highest] temps. Intel has a publicly available white paper on the what's and how's of it all if you want to dive in that rabbit hole.



There is the caveat that is that flag is being checked ONCE in 2 second intervals but has changed many millions of times in that 2 seconds.



The whole thing is much more complex than I addressed here so I did make some generalizations but it's safe to say that your PC and the previous commenter's are functioning nominally.



Is it trying to eat you?: No

Someone else?: No

Meh. You'll be fine then.
 
Back
Top