As the subject says, the thermals on my HX370 are kinda nuts. I found another thread that says the CPU will throttle at 100C. I feel it should only get to that level of heat doing something intense. Mine is sitting between 80-85C just downloading Fortnite. Not playing, just downloading.
This has been a thing with every Framework 13 I’ve owned and it’s starting to concern me. I have this issue with my Core Ultra FW13, I had it with my original 11th gen FW13.
My Core Ultra is running Mint so I can get where the OS may be part of the issue. My HX370 tho, is running Windows 11.
I just can’t figure out how to address the heating issues with the laptops. As I mentioned above, HWMonitor is showing 80-85C right now while downloading the latest Fortnite update. The power button and top part of the keyboard above the F keys and bottom bezel of the display is definitely more than warm to the touch.
The one thing that I do with all of them is some resolution scaling, primarily cuz I’m old. I have no idea if that would contribute to some of this. I’ve seen folks mention re-applying thermal paste or going with thermal pads instead. I love these laptops (I have 3 after all) but the heat aspect worries/bothers me.
For anyone who has had this issue, what things have you tried that provided the best result?
My Specs
CPU: AI 9 HX 370
RAM: 64GB
Storage: 4TB WD Black
OS: Windows 11
Repaste. Didn’t get that board, but one of the 7840U’s, and it ran hot until I repasted. I think your laptop should have PTM 7950 thermal pads already, so this is…concerning. I don’t know what would work better than that for your CPU.
Another thread had a poster with major thermal throttling issues; cause ended up being, at least in part, RAM chips with components on both sides of the chip. Some people engineered some solutions with thermal pads, but the issue was ultimately with the RAM. Might want to dig up that thread (I’ll see if I can find it myself) and take a look there.
Are all three of them roughly running close to max temps? I’d start with the 11th gen board for a repaste as that probably doesn’t have PTM. See if that works.
Believe it or not, “just” downloading a game can be a very cpu intensive task because they’re often decompressing on the fly. I don’t know about fortnight specifically, but steam for example is capable of loading 64+ cores while downloading on high bandwidth connections.
What are your temps like when you’re completely idle? You may not have anything to worry about - 85C is a completely normal temperature for a soc in a laptop chassis that’s doing substantial work. If it’s 85C while truly idle, that’s probably not so normal.
So I messed with things a bit more and yes, it appears the insane amount of horsepower needed to download and unpack Fortnite was definitely part of the problem. Once it cooled down it hovered around 65-70C. At least that explains one issue.
What worries me more with all of them is how hot the top of the keyboard deck, right under the display gets with pseudo-conventional work being done. I say “pseudo-conventional” cuz I figure most folks have only Outlook and a browser open where I would have more. My standard workflow would have Chrome, Edge, VSCode, Discord and Joplin open and that top can get pretty uncomfortable.
Overall I’m guessing the heat thing is gonna be relatively normal as I have had the issue across the board (no pun intended) with each one.
@MrTrismegistus - I wasn’t sure about PTM on the latest AMD so if they already have that, I don’t want to go down the route of replacing it if I don’t have to. And as far as memory, I stick with G.Skill and match speed to what’s required.
Hey @x12Mike glad to see that you identified the immediate issue. If it shows up again, give our support a shout here. They’ll be happy to help you narrow the culprit down further.
Yeah, because of on the fly decompression game downloads through clients tend to hit the CPU hard. Honestly though, 65-70 still sounds very high for a true idle to me. With basically nothing running I’d expect soc to fall to lowest p state, with package power maybe somewhere between 3 and 7 watts. At that power I’d expect temperatures to be somewhere around 15-20 c above ambient with fans off or spinning very slowly.
I wonder if maybe you still have something running that stops the soc from hitting lowest power state, or maybe a performance power profile that’s keeping the chip clocked up.
If you’re actually at lowest pstate and still seeing temperatures in the 65-70c range I might start worrying about cold plate mounting pressure and thermal paste application because that doesn’t sound normal to me - although I will say if you go from a loaded to unloaded state it might be reasonable to expect it to take ~5ish minutes for temperatures to fall.
Take what I say with a grain of salt though, I don’t have direct experience with the strix point chips, so I’m generalizing.