Power Limit Throttling

[DosxNova]

I've been facing this weird issue for a while now where the laptop is suffering from Power Limit Throttling from time to time. The cpu temps stays at max 80-86 and doesn't thermal throttle yet hits power limit throttling.

The CPU core clocks at 0.8 GHz while suffering and causes immediate crash on heavy apps, not to mention even Adobe apps crashes under relatively medium synthetic loads, I've tried resetting BIOS, reinstalling Win-10 again and again, also tried cleaning entire machine, been to ROG center to repair and yet it's still suffering this issue. In this pandemic I wasn't able to show it to a better repair center yet no one was able to narrow it down to any issue.

My config is Asus Strix GL502VM 

i7-7700HQ GTX 1060 3GB 8GB DDR4 2400MHz memory coupled with 500GB WD SSD and 1TB of HDD 

I've undervolted CPU a bit and tweaked with Core ICC and AVX Ratio recently but the results are same, worse the XTU score is as low as 512! More than half the score when it was functioning properly. 

I hope someone can help me, the only thing left is to reapply thermal paste by myself as the people at Asus did it already and it didn't work either, "The temps aren't even hitting 90 so it's not thermal issue", they said 

[samcool55]

Well the CPU does seem to run fine according to the graph, but as it says, its a power limit.

Everything is fine but the CPU is just not allowed to use more power. And with the power it can get, it performs whatever it can so with that limited amount.

 

Just reset everything to default and keep ONLY the core voltage offset. Lower voltage = lower power usage which (in theory) means it can go a bit faster (because it uses less power for the same performance).

[emosun]

5 minutes ago, DosxNova said:

The temps aren't even hitting 90 so it's not thermal issue

theres a big difference between thermal throttling and "how hot the cpu gets before it shuts the hardware off entirely"

 

Having your cpu in the 80s is hot enough to warrant thermal throttling to get the machine down to a reasonable temp. 

 

But purposely running the machine hotter than it should be and ignoring the temp while trying to place the blame elsewhere isnt going to fix the machine. Get the temps at least in the 70s

[DosxNova]

3 minutes ago, emosun said:

theres a big difference between thermal throttling and "how hot the cpu gets before it shuts the hardware off entirely"

 

Having your cpu in the 80s is hot enough to warrant thermal throttling to get the machine down to a reasonable temp. 

 

But purposely running the machine hotter than it should be and ignoring the temp while trying to place the blame elsewhere isnt going to fix the machine. Get the temps at least in the 70s

I actually haven't seen my laptop hitting any less than 90s and initially it didn't throttle even at 96-98, when i first brought it, using XTU i was able to undervolt it to lower the temps to 80 at max 75 (Like ever) even with ambient temps being 26-29 I'm not sure if only my machine is suffering this issue as when I looked up online this particular model showed better thermals compared to mine, but it never throttled up until now, it just popped out of nowhere. I'm not sure what triggered it as I didn't change anything in XTU prior to the appearance of this error. At load the temps even with undervolt stays at 76-78 (If it's not hitting Power Limit Throttling). I'm not sure anymore on what tweaks it's going to need to make it perform better than now. 

[emosun]

I cant imagine burning the cpu at 98c did the machine any favors. Just goes to show the result of ignoring temps. The hotter they run the sooner they die

[DosxNova]

10 minutes ago, samcool55 said:

Well the CPU does seem to run fine according to the graph, but as it says, its a power limit.

Everything is fine but the CPU is just not allowed to use more power. And with the power it can get, it performs whatever it can so with that limited amount.

 

Just reset everything to default and keep ONLY the core voltage offset. Lower voltage = lower power usage which (in theory) means it can go a bit faster (because it uses less power for the same performance).

Tried with no success, all default values and undervolt, n change in warning, without undervolt and default values, it hits thermal throttling first and power limit later! I never changed anything except Core Voltage Offset prior to the times since i'm getting these warnings. I'm at a point where i'm even confused about the default values in XTU. My machine always ran at high temps right out of the box, like 98 hot at loads, but never throttled and now it hits the mark at 80-82. I'm not sure if it's only my machine or what but it's been like that for past 6months and ROG center gave up on it as well. This new problem is killing so much of my time.

[DosxNova]

3 minutes ago, emosun said:

I cant imagine burning the cpu at 98c did the machine any favors. Just goes to show the result of ignoring temps. The hotter they run the sooner they die

I agree higher temps killing machine, but i'm saying I did what I could, the limit of  undervolting on this machine is 0.100V and I've been on it forever yet the temps won't go any lower than 80 (Even with Cooling pad). Other machines with same config are running cooler so I'm not sure if it's something I'm doing wrong or it's just the way this particular machine works (the cooling is surely terrible on this laptop I Agree). My primary workload is too CPU intensive so I'm quite concerned if the clock drops as the softwares keeps crashing. If it's gonna take more than XTU tweaks I'm willing to do it only if it works! 

[emosun]

11 minutes ago, DosxNova said:

If it's gonna take more than XTU tweaks I'm willing to do it only if it works! 

set affinity to less cores , but if that cpu is fried then idk what else you can do it with it. laptops aren't really made for heavy workloads thats what desktops are for. No sense in treating it like a desktop if it quite literally does not have the sheer mass of metal and cfm needed to run those components at full speed indefinitely.

if the temps just keep running away then it was the wrong machine for the job. 

[DosxNova]

Defaulting the values brought some numbers up but the warning is still up! 

[DosxNova]

3 minutes ago, emosun said:

set affinity to less cores , but if that cpu is fried then idk what else you can do it with it. laptops aren't really made for heavy workloads thats what desktops are for. No sense in treating it like a desktop if it quite literally does not have the sheer mass of metal and cfm needed to run those components at full speed indefinitely.

if the temps just keep running away then it was the wrong machine for the job. 

Is there any way I can verify is CPU is alright? If it's fried I think I'll need to switch soon! 

[emosun]

3 minutes ago, DosxNova said:

Is there any way I can verify is CPU is alright? If it's fried I think I'll need to switch soon! 

could stick the machine on a long time consuming render that'll take several hours , but doesn't actually load the cpu fully. Like timewarping an hour off footage or something. it'll at least confirm that it can be stable so long as you are loading the machine with a task it's designed to do rather than expecting it to go at 100% usage indefinitely which its definitely not designed to do.

[DosxNova]

4 minutes ago, emosun said:

could stick the machine on a long time consuming render that'll take several hours , but doesn't actually load the cpu fully. Like timewarping an hour off footage or something. it'll at least confirm that it can be stable so long as you are loading the machine with a task it's designed to do rather than expecting it to go at 100% usage indefinitely which its definitely not designed to do.

Thanks a lot, I'm quite happy if it doesn't crash in Adobe, i switched to desktop for renders and gaming long ago, I was having problem with Adobe recently that was my main concern. I stopped hammering it with heavy tasks a while ago. I hope it works! 

[LogicalDrm]

-> Moved to Troubleshooting