The thermal throttling out of the box is normal on these laptops. For the CPU anyways on Intel. Those readings also are ultra sensitive as to refresh every second or so, it may hit those temps for very briefly then go down but HWinfo still shows it's peak. Just saying don't over worry. Keep things on balanced mode, or good custom mode settings, proper ventilation 

But those CPUs are well known to be hot. 

Yes, it's throttling but it's also pulling nearly 110W...in a laptop. To put that into perspective the Ryzen 7 5800H in my '21 L5 Pro is rated for 45w with boosts up to 60-65w depending on the load and is just as hot at that point.

It's just the nature of modern high end laptops. Look into undervolting and power limiting. If you want a really quick and dirty solution, you could even lock the framerate in the Nvidia Control panel.

Oh, and just in case you've missed it. There are potential display and performance issues with the latest Nvidia driver, use 577.xx or older.

Running the 4070 with a I9 14900hx. Highly recommend undervolting. My cpu runs idle at a cool 60c... That being said I don't let it hit its full potential. Even on demanding games, but regardless better to have a working cpu than one capable of heating my home on a cold night. Outside of that make sure you raise it, I place a book under mine and the airflow makes a noticeable difference. As for the fans you're kind of out of luck there. The noise never bothered me, but I knew what I was getting into when I got a gaming laptop.

lmao it’s normal i got this computer last month and the fans stay at 4k rpm even just on marvel rivals or demanding games a on roblox even with a cooling pad you’ll get used to it 😭😭😭

Spikes are normal, if you are not found of them you can undervolt and limit the PL1 and PL2 via throttlestop.

They can be. I use mine flat on a table all the time and it does just fine. Raising the back is "ideal" but the temps stay controlled even without it (on my 7i gen 9 atleast) . If you look at the image op provided you'll notice the cpu is spiking to 110 watts and this was probably tested running a game so it's safe to assume the gpu was under load as well. These temps are normal in that case especially with lenovo's extremely conservative fan curves. If this was a cpu only benchmark then I'd be a bit concerned.

Same with my 5 Pro, thermal trottling after gaming for more than 30 mins

For further assistance be sure to visit us at our Discord Server to get help from our wonderful community!

I am a bot, and this action was performed automatically. Please contact the moderators of this subreddit if you have any questions or concerns.

Hi im having the exact same issue, did u ever find a solution?

Everything you've reported here is abnormal behavior. Send it back under warranty again.

There is likely some fault with powerdelivery, th CPU or the motherboard itself.

The 5800h can "only" clock down to 1.400mhz minimum, yours is at 400mhz min/max, which is unusable low.
Usualy this means that the motherboard, or BIOS, is currently running the system in some sort of safemode due to something being wrong.
On intel chips this is fairly common and its called PROCHOT. AMD chips are more locked down and behave differently ontop of it, but its very likely a similar core issue.
If you want to look into it you should use hwinfo64 instead but you wont be able to fix it as there is no great program like Throttlestop for AMD chips, so basically its RMA time.

You could try to reset the BIOS to default settings, but I highly doubt thats going to do anything.

Not reading all that but you can you throttlestop to disable turbo. That will drop temps by about 20c. It will also drop performance by a little bit but you can just turn off throttlestop when u need the full performance.

You shouldn't be disabling turbo unless you are happy with half the performance of what the laptop is capable of. Either replace the laptop or take the heatsink off and figure out why it's over heating so bad. For perspective, my strix can hold 195 watts in cinebench without thermal throttling.

Use ThrottleStop 9.7.3

Set all the turbo ratios to 52. This will allow you to reliably run a much bigger undervolt. My 14900HX is very stable with the CPU Core and P Cache set to -150 mV. V/F Point 1 should be set to 150. This was called mV Boost in older versions of ThrottleStop.

If you want to use the 58 multiplier when 1 or 2 cores are active, set the 4 highest MHz V/F points to 50. This will add some extra voltage where it is needed. Do some Cinebench single core testing to make sure the 58 multiplier is stable.

Do not reduce IccMax. This can cause instability. Set IccMax for the CPU Core, P Cache, Intel GPU and iGPU Unslice all to the max, 511.75. Also set Power Limit 4 to the max, 1023.

Do not check Speed Shift EPP. Allow the Windows 11 power plans to control EPP. Setting EPP in ThrottleStop can cause a fight between ThrottleStop and Windows over control of EPP. It is always best to avoid instability.

The 14900HX has amazing potential with good cooling when it is properly setup.

https://i.imgur.com/jAG9ZsJ.png

If your computer is too hot, lower the turbo power limits. Somewhere around 100W to 140W for PL1 and PL2 should be a good compromise when playing games.

Simply repaste the original Liquid Metal. It works really well - After that my fan never goes over 3500 rpm and keeps under 90c

I have the same machine as you, and just encountered the exact same issue. I opened my machine a few days ago to find the center of the CPU die almost completely bare of liquid metal. I am currently waiting on new liquid metal and PTM7950 to arrive, and I'll let you know how the self-repair process goes.

Note that I had to carefully remove the tamper-evident sticker on the top right screw to the thermal assembly during user-repair. I find it unacceptable that users aren't allowed to fix their own devices under warranty even though this issue is so common.

Attaching picture of replaced heatsink, you can clearly see that print is not on all CPU surface.

https://preview.redd.it/jytr4kg9iaue1.jpeg?width=1576&format=pjpg&auto=webp&s=be267c170869e933b9caf1328b998797e6618d17

> I noticed that the liquid metal application on the CPU was uneven, covering only about half of the die

How do you even check this?

i have the same one bought about two weeks ago , i also have the same issues. It reaches 95C and thermal throttles in Cinebench R23 single core tests. Even when booting up it reaches 100C, idle temps are also 70 -80C. Definitely not normal. I'm also trying to get a replacement.

I have the same model. Before undervolting it, R23 scores (ran for 10 mins): Multi core-21698 Single core-1959 After Undervolt :- Multi core-24673 Single core- 1993 Room temperature was around 10*C (I did this test last winter around January). Haven't tested this summer but your scores are way less. I guess go for a replacement if possible

The important thing is what are temperatures like when you are playing AAA games at maxed setting? Play something for 30min and check HWINFO.

Wtf happened to your CPU cores? 14650HX is a 16 core CPU, yours only has 5

I had the exact same model with the same cpu. I think its an issue with a lot of 14650hx cpus. Mine went to 105 at times peaked there and throttled. And my unit was defective after 2 weeks in which time i got a refund. Ive now purchased the legion 5i gen 10 i hope it doesnt have the same issue

I don't think it is a thermal throttle. I think you should updatethe Intel Management Engine firmware (don't forget to run it as administrator) via lenovo support (manual)the biosthe energy driversReinstall the thermal driver manually

how long have you had this before you ran into thermal throttling? from day 1 or a problem that crept over ? And how is your laptop now?

If you read some of the reviews on Notebookcheck, it seems that Intel Gen 12 and Gen 13 P and H CPUs have throttling issues in many devices. Those processors are simply not very power efficient. They still have to use a quite large stepping - they just renamed their manufacturing process so it sounds better. In those tests I mentioned above you will see that most (all?) of those Intel based machines throttle after a short time. As long as you don't have performance issues, I wouldn't bother. But I would not try to push the CPU even harder. Don't run it aggressive, don't overclock it.

Try using a cooling pad

Facing the exact same issue! Did you resolve it?

Hi, guys. First of all, thank you so much for your comments.

I did not check the temps before update do Windows 11, because I litteraly did not used the laptop. Updating to Windows 11 was one of the first thing that I did. After my post here, I played Halo infinite for around 20 minutes, and when I checked the temp after exiting the game, the max hit was 90ºC + and the thermal throttling was marked as yes in the HWinfo. I also noticed some little stuttering during the gameplay. However, I did not check the clock, temp and CPU% usage at real time.

In this morning I decided to open up the back and to a repaste. That was what I found when I took the heat sink:

https://imgur.com/a/qF51ihU

One termal pads was misaligned misaligned place badly. I managed to try to get it back to the right place.

https://imgur.com/a/Sd5TSCS

I also did a repaste. The stock thermal paste was a bit hard and dry.

After that, I played Halo Infinite again, for around 10 minutes and my temps were way better.

I have an average around 55 - 60Cº. But the peak was 85º. And the best news is that the thermal throttling wasn't marked as yes in the HWinfo this time.

So I think the problem was the thermal pads and the bad thermal paste. Probably the product was assembled by Lenovo some time ago which mad the paste get dried.

What's the update? How's your PC running? I have a 5i as well with a basket of issues.

​

EDIT: Nevermind saw the post about you re-applying the thermal paste. Will let you know how it goes. This laptop completely shits the bed on any game it plays I assume that's why. I'm beginning to believe a lot of lenovo legions come with poorly applied paste.

Sounds strange. My legion 7i doesn't go last 65 -70 even when playing games at max settings.

This is definitely fishy... The CPU is not supposed to go to those temps when running OS updates or downloading/installing games and/or apps. Unless you had another CPU intensive process running at the time?

I ran into this issue as well. I changed processor power management to 92% per Lenovo support and seem to fixed it. Hope this helps someone with core(s) throttle issues in both performance and balance setting.

Yeah something is definitely off. If you hit 100 degrees your cpu/GPU should throttle. That’s how they are designed. Do you have a cooling device underneath it ? Most people on here recommend that. The Lenovo techs are probably correct in that the thermals are working correctly because they are throttling as intended. You just need to figure out how to get your temps down a bit. It’s weird that you are reaching high enough temps to throttle in quiet mode. Check your temps at idle and see what they look like.

Same laptop here. If your CPU gets hot, but gameplay does not suffer, then it’s working as intended. While playing cyberpunk, I get 92 C average temp with spikes going up to 95 C and sometimes even 100. It’ll just touch it and come down and be stable at 95 .. this is in auto thermal profile.

Gameplay is smooth as butter for me.

In silent mode clock speeds are reduced so temps are down too. For the same game I get 78-85 C on silent mode. Performance on gameplay is fine but it stutters when it has to auto save . It’s very momentary and can be ignored but I’m just gonna play on Auto mode. I want a great visual experience. Which is my reason to by this laptop.

It should not hit 80 C while in idle, however sometimes background apps like windows defender can cause the fans to spin up. Or temps to increase while you are idle.

Actual idle temps are between 45-60 C for me.

I’d suggest you operate in silent thermal profile for some time, monitor temps with Hw info, monitor performance too.

Then shift to “balanced and turn on Lenovo AI engine” Let it learn your daily use case. Don’t get anxious bout temps unless there’s performance issues.

Faulty vapor chamber

There's something wrong with your laptop. I mean, this model is hot but 80 degrees on idle - is not normal. I had about 38-45 (before my LM experiments). 

I would try to repaste it myself using PTM7950 and get a proper laptop cooler. I recommend FlyDigi BS1 (the most quiet on the market right now). I have the same laptop and with this cooler on medium speed (not loud) CPU usually stayed at 70-85 in AAA games (performance mode). Then, I replaced PTM7950 with liquid metal for CPU and now it stays at 68-73 during gaming and the laptop is much much quieter now.  GPU isn't hot on this laptop so PTM7950 works perfectly fine for it.

Even without the cooler and LM it never throttled, never went over 92 degrees as I remember. And I live in Cyprus. It's freaking hot here. So if it throttles with PTM7950, you should RMA.  Also I heard that Lenovo technicians can use regular thermal paste after their repairs (this was my case, unfortunately) so you better repaste it yourself. If you do so, don't use anything except PTM7950. It's the only proper thermal material for this laptop. You won't get better temps with any another paste. 

(Sorry for my English, it's not my native language)

Tech probably used normal paste which would pump out.

I mean its hard to judge with techs, every pro I've had work on my machine has been sub par and I've done better myself.

For it to randomly get really bad, if it does have a vapour chamber, it may have failed? Otherwise id be stripping it down, taking the fans off and deep cleaning those and the stacks. Polishing contacts points, thoroughly cleaning with Isopropyl 99%. Me personally, ai'd be liquid metal'ing that 5900HX, fresh PTM on the 3080 and UTP-8 and the VRAM and VRMs.

Though, if that VC has failed... it wont help you much.

Does it have a performance mode that sets the fans to max?

A repaste of the GPU and CPU, along with cleaning out the fans, would probably do more good than undervolting at this point.