GE76 CPU 12th Gen Temperatures

qaziahme159302de said:
waiting for the weekend so I can do a re-paste and do much more including a full 10 minutes test :p
Click to expand...
If you want extreme jump, buy liquid metal paste ;). My jaw dropped. I removed at first thermal phase change paste (i thought i was stupid). I start with grizzly kryonaut. The results were worse, so i thought, that ill cry. :D Second paste was grizzly kryonaut extreme. With this paste i almost equal to the origin results. In same way i thought that they were better, but it doesnt. After days of use, the results were worse again, so last chance was grizzly conductonaut and this is the big f.....g win. In all cases i did some undervolting P-Cores -115mv E-Cores -50mv Ring(Cache) -47mv. In full load i wasnt able get more than tdp 70w. It lavitates between 65-70w, after applying liquid metal, im all the time between 95w - 105w. Whaaaat? Im so exicited. :D Finally i found cure on my soul. All the time im talking about CPU. Regarding to GPU, i think that i didnt apply liquid metal correctly, but im not sure about that. Maybe not. If not, it is not that big jump as CPU, but result are good i think anyway, so i didnt do repasting. On TDP 150W (extreme performance profile) it hits 78 degrees after 10minutes furemark (+200mhz gps +200mhz memory). What is better before grizzly kryonaut extreme is, that i can play on silent mode without hitting thermal throttling. silent mode is 2800 RPM, but max GPU TDP is 125W. It hits 81 degree. I have better result but i expected more from liquid metal. This is why think i didnt apply it correctly, but they are better, so it will wait till next time repasting. :D Today i finally have Christmas. Next thing with what ill play is hitting to 5GHZ on single core. What i saw from chart, maximum value was 4.85Ghz, but very shortly, avg. is around 4.7Ghz. I dont know why. This is what i got from one month tuning with 3200eur laptop with 12900hk and 3070ti. What i can say. Never give up brother. :D

PS: cinebench on cold pc, one round 19600 pts ;) this is reason, why it is not relevant
PS2: what next i got from liquid metal on CPU is stable rising temperature based on load. On previous pastes it jumps from 65 to 95 on 20% load. Now is stable around 60 degree and it is quiet (2800rpm) instead of hitting 4000 - 5000 rpm on 20% load.
PS3: i use cooling pad iets gt500 v2, but on lowest possible speed because it is too loud on higher rpm.

1647374710333.png
 
Last edited:
amzuk.leina158702e0 said:
If you want extreme jump, buy liquid metal paste ;). My jaw dropped. I removed at first thermal phase change paste (i thought i was stupid). I start with grizzly kryonaut. The results were worse, so i thought, that ill cry. :D Second paste was grizzly kryonaut extreme. With this paste i almost equal to the origin results. In same way i thought that they were better, but it doesnt. After days of use, the results were worse again, so last chance was grizzly conductonaut and this is the big f.....g win. In all cases i did some undervolting P-Cores -115mv E-Cores -50mv Ring(Cache) -47mv. In full load i wasnt able get more than tdp 70w. It lavitates between 65-70w, after applying liquid metal, im all the time between 95w - 105w. Whaaaat? Im so exicited. :D Finally i found cure on my soul. All the time im talking about CPU. Regarding to GPU, i think that i didnt apply liquid metal correctly, but im not sure about that. Maybe not. If not, it is not that big jump as CPU, but result are good i think anyway, so i didnt do repasting. On TDP 150W (extreme performance profile) it hits 78 degrees after 10minutes furemark (+200mhz gps +200mhz memory). What is better before grizzly kryonaut extreme is, that i can play on silent mode without hitting thermal throttling. silent mode is 2800 RPM, but max GPU TDP is 125W. It hits 81 degree. I have better result but i expected more from liquid metal. This is why think i didnt apply it correctly, but they are better, so it will wait till next time repasting. :D Today i finally have Christmas. Next thing with what ill play is hitting to 5GHZ on single core. What i saw from chart, maximum value was 4.85Ghz, but very shortly, avg. is around 4.7Ghz. I dont know why. This is what i got from one month tuning with 3200eur laptop with 12900hk and 3070ti. What i can say. Never give up brother. :D

PS: cinebench on cold pc, one round 19600 pts ;) this is reason, why it is not relevant
PS2: what next i got from liquid metal on CPU is stable rising temperature based on load. On previous pastes it jumps from 65 to 95 on 20% load. Now is stable around 60 degree and it is quiet (2800rpm) instead of hitting 4000 - 5000 rpm on 20% load.
PS3: i use cooling pad iets gt500 v2, but on lowest possible speed because it is too loud on higher rpm.

View attachment 157842
Click to expand...
Good for you! Though gotta be careful with conductive paste when applying. Also, check the longevity of the paste because i’ve read somewhere that liquid metal lasts less long compared to non-conductive. Just reapply in that case.
 
amzuk.leina158702e0 said:
If you want extreme jump, buy liquid metal paste ;). My jaw dropped. I removed at first thermal phase change paste (i thought i was stupid). I start with grizzly kryonaut. The results were worse, so i thought, that ill cry. :D Second paste was grizzly kryonaut extreme. With this paste i almost equal to the origin results. In same way i thought that they were better, but it doesnt. After days of use, the results were worse again, so last chance was grizzly conductonaut and this is the big f.....g win. In all cases i did some undervolting P-Cores -115mv E-Cores -50mv Ring(Cache) -47mv. In full load i wasnt able get more than tdp 70w. It lavitates between 65-70w, after applying liquid metal, im all the time between 95w - 105w. Whaaaat? Im so exicited. :D Finally i found cure on my soul. All the time im talking about CPU. Regarding to GPU, i think that i didnt apply liquid metal correctly, but im not sure about that. Maybe not. If not, it is not that big jump as CPU, but result are good i think anyway, so i didnt do repasting. On TDP 150W (extreme performance profile) it hits 78 degrees after 10minutes furemark (+200mhz gps +200mhz memory). What is better before grizzly kryonaut extreme is, that i can play on silent mode without hitting thermal throttling. silent mode is 2800 RPM, but max GPU TDP is 125W. It hits 81 degree. I have better result but i expected more from liquid metal. This is why think i didnt apply it correctly, but they are better, so it will wait till next time repasting. :D Today i finally have Christmas. Next thing with what ill play is hitting to 5GHZ on single core. What i saw from chart, maximum value was 4.85Ghz, but very shortly, avg. is around 4.7Ghz. I dont know why. This is what i got from one month tuning with 3200eur laptop with 12900hk and 3070ti. What i can say. Never give up brother. :D

PS: cinebench on cold pc, one round 19600 pts ;) this is reason, why it is not relevant
PS2: what next i got from liquid metal on CPU is stable rising temperature based on load. On previous pastes it jumps from 65 to 95 on 20% load. Now is stable around 60 degree and it is quiet (2800rpm) instead of hitting 4000 - 5000 rpm on 20% load.
PS3: i use cooling pad iets gt500 v2, but on lowest possible speed because it is too loud on higher rpm.

View attachment 157842
Click to expand...

not in favour of liquid metal due to long term heat sink corrosion and the danger of it spilling out

12900HK comes with a border around the CPU for liquid metal safety...did you remove it? if yes then how are you making sure that liquid metal doesn't spill around?

I am planning to put grizzly kryonaut extreme on CPU/GPU and k5 pro around
 
akuzacj158a02e3 said:
does changing the paste or even opening the laptop void the warranty?
Click to expand...
No it doesn't, except if you break it in the progress obviously. However, what people here are doing is utterly pointless. Original Paste is absolutely fine and sufficient, just use Cooler Booster for hardcore gaming and/or get a cooling pad. The Laptop has way more power than it should have considering how thin it is, no paste in the world will change anything about your CPU capping 90 degrees and throttling during gaming.
 
qaziahme159302de said:
not in favour of liquid metal due to long term heat sink corrosion and the danger of it spilling out

12900HK comes with a border around the CPU for liquid metal safety...did you remove it? if yes then how are you making sure that liquid metal doesn't spill around?

I am planning to put grizzly kryonaut extreme on CPU/GPU and k5 pro around
Click to expand...

yes i removed it, one of my stupidity, what i done. How do i know? Im writting from the laptop, so even if yes, it is not something dangerous for laptop for now ;). On GPU i covered with classic paste place which could be sensitive for touching with liquid metal, but on cpu is nothing and from photo what i did for this thread https://forum-en.msi.com/index.php?threads/12900hk-experiences.372578/post-2111976 there is lot of safe place around.

PS: copper is fine with liquid metal in all reviews what i saw, it changes color, but there are no damages.
 
Last edited:
bat-l15b602e7 said:
No it doesn't, except if you break it in the progress obviously. However, what people here are doing is utterly pointless. Original Paste is absolutely fine and sufficient, just use Cooler Booster for hardcore gaming and/or get a cooling pad. The Laptop has way more power than it should have considering how thin it is, no paste in the world will change anything about your CPU capping 90 degrees and throttling during gaming.
Click to expand...
i bought this laptop not for gaming, but for programming and the best cpu is just in gaming laptops and i need cpu performance on first place. And jump from 70w in full load to 100w, it makes mi point to do that, but you are right, that it is not necessary if you change it for another paste, but liquid metal is another level. Im talking about 12900hk which has thermal phase change paste. I cant say for 12700H laptops, somebody mentioned, that there is classic paste.
 
bat-l15b602e7 said:
No it doesn't, except if you break it in the progress obviously. However, what people here are doing is utterly pointless. Original Paste is absolutely fine and sufficient, just use Cooler Booster for hardcore gaming and/or get a cooling pad. The Laptop has way more power than it should have considering how thin it is, no paste in the world will change anything about your CPU capping 90 degrees and throttling during gaming.
Click to expand...

yeah i was wondering that actually, right now the GE76 with i7 +3070ti or GE66 i9 3070ti (both are available where I live) are my favourites to get on early april so wonder if i should be repasting myself, if most likely my only game this year will be elden ring, but wouldnt want my laptop to run too hot/loud, around 85-90 max i guess

but then isnt the i9 already with thermal liquid? are they terrible and its better off to repaste with grizzly conductonaut ??
 
pacheco.phili151502d4 said:
Good for you! Though gotta be careful with conductive paste when applying. Also, check the longevity of the paste because i’ve read somewhere that liquid metal lasts less long compared to non-conductive. Just reapply in that case.
Click to expand...
yep, i count with that. I think each 2 years it will be necessary. But ill see result after year, if there is big temp degradation.
 
amzuk.leina158702e0 said:
i bought this laptop not for gaming, but for programming and the best cpu is just in gaming laptops and i need cpu performance on first place. And jump from 70w in full load to 100w, it makes mi point to do that, but you are right, that it is not necessary if you change it for another paste, but liquid metal is another level. Im talking about 12900hk which has thermal phase change paste. I cant say for 12700H laptops, somebody mentioned, that there is classic paste.
Click to expand...

How would you need a 12700H running with 100W need for programming? That makes no sense. Maybe for compiling, but then again, why wouldn't you just get a Workstation for that and just use some average office laptop for the actual programing?

And I think liquid metal ruins more than it does good, especially if the device wasn't designed to be used with liquid metal. dont want anyone here opening their laptops to break it by trying to apply liquid metal. no other normal paste will do better than the pre-installed paste.

also, with cooler booster and without cooling pad, my laptop still hits up to 90W:
1647419754145.png

Which it is capped on by me via Bios. So yeah, I still doubt that all that tinkering with the laptop actually provides you more than a placebo effect.
 
akuzacj158a02e3 said:
yeah i was wondering that actually, right now the GE76 with i7 +3070ti or GE66 i9 3070ti (both are available where I live) are my favourites to get on early april so wonder if i should be repasting myself, if most likely my only game this year will be elden ring, but wouldnt want my laptop to run too hot/loud, around 85-90 max i guess

but then isnt the i9 already with thermal liquid? are they terrible and its better off to repaste with grizzly conductonaut ??
Click to expand...

If doesnt matter size for you, GE76 is better choice, better cooling. i have g66 i9 and the cooling is not good enough with their thermal phase change paste either. If you check my post above, there is just one way to get good performance from smaller case, liquid metal, with this CPU.
 
Last edited:
bat-l15b602e7 said:
How would you need a 12700H running with 100W need for programming? That makes no sense. Maybe for compiling, but then again, why wouldn't you just get a Workstation for that and just use some average office laptop for the actual programing?

And I think liquid metal ruins more than it does good, especially if the device wasn't designed to be used with liquid metal. dont want anyone here opening their laptops to break it by trying to apply liquid metal. no other normal paste will do better than the pre-installed paste.

also, with cooler booster and without cooling pad, my laptop still hits up to 90W:
View attachment 157865
Which it is capped on by me via Bios. So yeah, I still doubt that all that tinkering with the laptop actually provides you more than a placebo effect.
Click to expand...
Exactly, compiling is that case and i have 12900hk in my case

with cooler booster? ok, so ill tell my colleges, that they have bring earplugs with them to work :D this is not solution and it is not placebo effect. I went through 3 pastes and one month tuning the laptop, so please dont tell me what is right and what is wrong. I share my experiences, maybe they will be useful for someone and i dont need run cooler booster to hit 100W. Just regular fan speed and in single thread full load is enough 2800rpm (60 degrees), can you say the same? yes disassembly is risk for user, so for that users, im saying, you need some experiences to do that. If you are regular user, dont do that.

you hit me with the cooler booster :D :D im still laughting.
 
amzuk.leina158702e0 said:
I went through 3 pastes and one month tuning the laptop, so please dont tell me what is right and what is wrong.
Click to expand...

Yeah, thats what I am saying. Placebo Effect. Putting so much worth into completely worthless repasting must have had a positive effect in your brain, otherwise you wouldn't be able to justify what you did to waste your time and potentially damage hardware.

without cooler booster, you wont be able to hit 100W anyway. No matter how well you transport the heat from the CPU to your block, you then need to get the heat out of the system. This can ONLY be done with air.

As I said, you need a workstation or a server, not a gaming laptop. So you basically have the wrong device and you try to find a solution for a problem you shouldn't have in the first place.
 
amzuk.leina158702e0 said:
Exactly, compiling is that case and i have 12900hk in my case

with cooler booster? ok, so ill tell my colleges, that they have bring earplugs with them to work :D this is not solution and it is not placebo effect. I went through 3 pastes and one month tuning the laptop, so please dont tell me what is right and what is wrong. I share my experiences, maybe they will be useful for someone and i dont need run cooler booster to hit 100W. Just regular fan speed and in single thread full load is enough 2800rpm (60 degrees), can you say the same? yes disassembly is risk for user, so for that users, im saying, you need some experiences to do that. If you are regular user, dont do that.

you hit me with the cooler booster :D :D im still laughting.
Click to expand...
yeah im leaning towards the GE76 however I am not sure if its screen has DCI P3 which is pretty much mandatory after using a macbook, but so far nobody could answer my question on the forums about it...
 
bat-l15b602e7 said:
Yeah, thats what I am saying. Placebo Effect. Putting so much worth into completely worthless repasting must have had a positive effect in your brain, otherwise you wouldn't be able to justify what you did to waste your time and potentially damage hardware.

without cooler booster, you wont be able to hit 100W anyway. No matter how well you transport the heat from the CPU to your block, you then need to get the heat out of the system. This can ONLY be done with air.

As I said, you need a workstation or a server, not a gaming laptop. So you basically have the wrong device and you try to find a solution for a problem you shouldn't have in the first place.
Click to expand...

i found solution, im still dont get your point. I reached what i need and you are not able accept it? Im satisfied, i get what i expected from this laptop finally. What placebo effect? I bought ge66 with i9 because it is the best mobile solution and the case can handle temps for such as cpu. Only way how unlock the full performance in that small case is liquid metal. This info is very important for people who need mobile platform in this small case from this manufacture. I didnt do any damage for my self, so what are you talking about? My case, my experiences, my results. Im sharing the info. Wtf man. What is wrong with you? :) Where is the problem? Still dont get it.

PS: 20% better time on compiling time (compare without liquid metal and with it), just for info, but i understand, it is not important for you, but for me yes. :) Less time, bigger efficiency for the client. I can do more work and not wait for compiling so long. This is how i approach to work and this is my goal, what i want reach with buying this laptop. Of course i couldnt be satisfied with result when cpu goes on 60%. I want 100% of performance and i need mobile solution and i did it, with upgrade and risk, but i did it.
 
Back
Top