Overheating and fan control issues with Predator Helios Neo 16 (PHN16-72). Looking for solutions?

Interesting how manipulative and half-lying your reply is. So let's break it down.

* Undervolting does NOT damage the cpu, by any means. By all intents and purposes, undervolting can actually PREVENT a cpu from getting damaged, and prevent it from degradation. Quite important ESPECIALLY since I'm talking about a 14900HX, one of the cpus which historically has had degradation issues. Also fun how you ignored the other comments where I showed the cpu going to 1.5v and above, which is actually damaging to the cpu, unlike undervolting. Any article online will show that undervolting does not and cannot damage a cpu. Also discussed and explained thoroughly by me in previous comments in this thread.

*Don't care about warranty. My warranty is null and void anyway, I replaced the joke "liquid metal" that Acer spat on that cpu, in order to attepmt to fix or at least lower temps. Which I did succeed at, but overall insufficiently, as the cpu still gives itself 1.45v and above, in its attempts to fry itself. A well known issue of the 13th and 14th gen flagship gpus.
Same can be said about the undervolt; it could have been allowed and just once activated, to void warranty (there are software fuses for this type of thing).
Furthermore, undervolting was allowed and then removed by Acer on this laptop model. Implying previous owners of this model were able to do this before Acer decided to take the option away from them. I would gladly do it at my own risk, but it's LOCKED. That's my whole point to begin with, stated from the start.

* Oh, really? Your I5-13500hx and rtx 4050 does not overheat? Thats nice. Also nice how you're using your lowest end cpu and gpu to compare to the 14900hx (famous for generating more heat, using more power, etc) and a far higher tier gpu.
Here's the fun bit of actual information: while the laptops have different spec, the model has the same cooling; that's how they're all produced, with only motherboard changes, but otherwise, keeping as many of the other parts standardized within a laptop model, to lower production and tooling costs. Your i5 + rtx 4050 has the same cooling elements as my i9 + 4070, while having almost half the number of cores/threads, and a far lower tdp/tgp. Don't even get me started on rtx 4050 vs rtx 4070. While having wildly different power draws, and you admitting to still requiring to use multi fan cooler on it. This shows just how misleading you attempt to be, writing the narative, sounding confident and giving "explanations" to defend Acer while comparing a bicycle to a muscle car.

* Cpu temps and different triggers for Tjmax, eh? There's a lot of asterisks to that chart that you forgot to mention, mainly regarding the actual performance that is gained/lost in all of them. The fact that the Predator Sense software forces the user to a specific minimum rpm when using a profile, or how it completely ignores user rpm setting once it reaches 80-82c (also a joke, they couldn't even implement an actual fan curve? really?) and the fact that each profile limits cpu tdp and gpu tgp, making the performance in between them be wildly different. And in none of them does this 14900hx+rtx4070 achieve even remotely the performance it should, due to throttling. All these which I've also detailed in previous comments.

If you would have read this whole post and my other replies here, you'd have noticed I'm far from being a beginner at pc hardware. Lying and trying to mislead me will not work, it'll only show what Acer themselves also showed me in their "support" messages: they chose to lie and spew bull than to actually provide their customers with a properly functional product.
Overall, your reply sounds like damage control, but sloppy work, with insufficient knowledge to actually argument against the points made in here. So instead, you compared lowest sku vs highest sku (well, except gpu, that's a mid-tier sku in mine), to claim that yours is fine, and then justify things under the guise of safety and warranty, although that's easily proven wrong by the simplest google search.

Hope Acer pays you well for your "work".

It is not harsh, it is realistic. If you saw it as insulting, that is your perspective.

You came in talking like an expert, finding dismissals for everything, without bothering to look through the whole thread for the fact that all those were either dismissed or explained as to why it does not work that way (or ignoring that altogether).

Your example of using the bottom of the barrel i5, mixed with the lowest tier and capability gpu from that generation, and then claim "you have no problems", so it's my fault or I'm exaggerating or it's my fault? Yeah. That is insulting. Especially when I clearly mentioned the fact that I'm talking about the i9 14900hx, explained its issues, it's potential degradation in time, and even listed multiple thermal interfaces attempted. Rtx4050 vs rtx 4070 is also a whole different game, and both cpu and gpu share the same heat spreader. Your 13500hx+4050 generates half (or less) of the heat that a 14900hx+4070 does, while running on the exact same cooling elements and fan setup.

You don't know those temps, since your cpu isnt even remotely in the same category of power consumption, clock frequencies, overvoltage from factory and heat generation. Not even close, contrary to your belief.

And those games are all old and optimized/low requirements enough that they are playable with ease on handheld pcs with far weaker specs.

You proceeded to upload a pointless fan setting printscreen from the manual, ignoring that my irks regarding the fans were related to the lack of a proper fan curve and complete lack of user setup, since it ignores whatever you set in Predator Sense as soon as it reaches 80-82celsius, moment in which the fans go ham causing excessively loud output.

I can keep it below Tjmax, especially after replacing the subpar "liquid metal" that Acer put on it with higher quality TMs. The issue I mentioned repeatedly was noise and lack of control over said noise.

As for profiles, they don't affect just fan speeds, they also affect clock frequencies of both cpu and gpu and their tdp/tgp.

You dismissed everything and went on tangents to find excuses, without even bothering to look through and see what has been discussed prior.

As for your "if you are so knowledgeable, find a bios program to undervolt". Yeah, no.

I specifically mentioned I'm a hardware guy. And even in that situation, I can tell you it would be a bios mod, custom mod or hack, not a "bios software".

And I wish that was possible, but Acer made sure to lock everything down to an extreme level. Although, as I said before, bios undervolt WAS indeed available at start on the Helios 16, and removed later on by Acer, with the provably wrong statement of "it damages cpu". No it does not. No case on the whole internet where undervolting harmed a component. It is a recommended tweak for cpus and gpus for decades. Worst case scenario, insufficient voltage causes instability. Bios reset, try again. Simple.

Twist it as you wish, what you did, how you tried dismissing and acting like an expert when you were so far off beat from the subject only makes me believe that you're a Acer defender, without being able to acknowledge the flaws that they have.

And it s a shame, too. This laptop series has great potential. It is dropped massively by such lockdowns chosen by Acer.

P.S. : Asus, Gigabyte, Msi, etc laptops of similar spec? Yeah, they allow undervolting.

I did not take it on "you guys".
I replied directly to a person that dismissed, ignored and excused a giant chunk of a large thread , in order to excuse Acer, and stated that since his 13500hx doesn't overheat that much, neither should my 14900hx. That alone, ignoring the rest of the things he said, was enough to warrant the reply he received. No problem if one doesn't have enough knowledge, none of us were born knowing things, BUT don't attempt to formulate a statement like it's pure fact when it's so far off, it's barely in the same zip code with the discussion at hand.
And again, it was selective response to one person, not a generalization towards a larger number of people. Direct reply, direct quotation.

Regarding your linked datasheet: While Intel recognizes Operating Voltage of a CPU as 1.72v maximum, that is an extreme case, and generally only ever gotten there through exotic cooling (LN2, Helium, etc. I've played with those as well in the past, on desktop parts), and guaranteed degradation.
Based on everything that has happened in the last 2 years, conclusions were drawn regarding this; first and foremost, the issue with the 13th and 14th gen, especially on the 13900/14900 SKUs, that caused the whole drama, was due to the cpu receiving too much voltage. The cache voltage being linked to vcore caused it to receive more than required, damaging it. It has been acknowledged that any voltage above 1.45v is almost a guarantee of cpu degradation.
I have personally went above 1.5v (1.529v to be precise, in the printscreen posted in this thread, 1.55ish volts in other cases when I didn't bother print screening anymore) in light usage of this laptop, even posted a printscreen of that earlier in this post, in a far older comment.
Even ignoring the potential degradation this can lead to, that voltage alone explains the overheat issues that I and others have complained about. Undervolting it, keeping it below 1.4v at all times, be it with a static voltage, or with an offset, would both guarantee a longer lifespan of said cpu (fewer RMA requests towards Acer), and also visibly, tangibly improve temperatures (fewer complaints towards Acer).
By all means, it is a win-win situation.
For perspective, my main pc has a 13900KF from launch day, which I've made sure to never touch 1.4v , let alone more than that. And after all these years, it is still running fine, contrary to what the media keeps saying about all 13900 and 14900 cpus. Built plenty of those two SKUs for customers, manually undervolted each one before giving it to their new owners, and all are still functional.

As for the "It was Intel itself that removed the voltage control (undervolt) on mobile 10th gen CPUs and the latest H series" statement:
They've removed the option being on by default. Any and all manufacturers can simply add it into their bios still.
As an example, a relative equivalent of this laptop from Asus, the Strix G18 (13980hx + rtx4070, 2024 model) I can guarantee you it does indeed allow undervolting, as I've done it for someone. Google will probably confirm this as well. I also undervolted some Msi laptops, but I forgot the exact models; remembered the Strix because it was more recent.


As for microcode and CSME updates improving it… permit me to be skeptical. Skeptical on Intel to do it properly, as they never have in this regard, and skeptical on motherboard/laptop manufacturers to tweak it to make it better.
Undervolting has been one of the main recomendations for any cpu of any company from the past 2 decades, and it has been such for a reason; from lowering temps to improving lifespan, it's just a good thing to do, and it's been clear, again, for 2 decades, that Intel and the manufacturers don't care or don't want to bother.
As for Acer's goodwill… I doubt that even exists. The feature was available when the Helios 16 launched, and then removed because they deemed it so. That's kind of the opposite of goodwill.


As for Plundervolt. Read your own link thoroughly. It implies they had to undervolt something massively, beyond the point of loosing stability, in order to use that as a backdoor.
It can be easily mitigated by not allowing apps such as ThrottleStop and Intel XTU to undervolt within the OS, and only allow undervolt done through Bios.
Many modern motherboards actually have a toggle in bios, if they are to allow OS-based apps to undervolt or not.

Very small "improvement" in VID when setting PL1 and PL2 to be 50w or below.
All the other, voltage-related options are locked (thus the existence of this post), so I cannot edit them.
All I can change in ThrottleStop , basically, is PL1, PL2 and Turbo time limit.

Tried both e-cores on or off (from bios), difference was close to non-existent in terms of temps and voltage. Changing their performance, I assume you mean frequency, that's locked, as everything else…

As for running only one nvme drive, that I have not tried. Given how the laptop has only two slots, while the 14900hx is capable of handling 3-4 nvmes in higher end laptops, I doubt that to be the case. 14900hx has 20 pcie lanes, gpu uses only 8 lanes, leaving the cpu with availability to sustain at least 3 nvme drives

The fact that the drive is connected to the cpu is doubtfully at fault, simply because in idle, the clocks themselves do scale down visibly, and there's barely any load on it whatsoever (according to Hwinfo, Task Manager, etc)
Might try the one nvme hypothesis at some point, in the weekend, when I can waste a bit more time, so I'll also disassemble the heatsink, re-reapply TG Putty Pro and ptm7950 to it, and take a picture of the whole thing for you, in the process. Thinking of replacing the OG drive that came with the laptop with a 4tb one, anyway; might as well do a test before swapping it.

My theory remains at VCC_Core being shyte and ignorable. Similar issue with my main pc, which is running a 13900KF since it launched. According to VID, it wanted 1.5v+, and it overheated like crazy. Yet it's been running for 3 years with a cap at 1.36v like a champ, outperforming the unlocked voltage by quite a margin in benchmarks and stress tests, while being a lot less… melty.

Similar situations for multiple other 13900/14900 builds I've made in the past, all wanted well above 1.4v by default, yet ran better and cooler with below 1.37v (silicon lottery, go figure. Varied between 1.3v and 1.37v,depending on chip luck). On the laptop side, the Strix G18 mentioned earlier accepted a -40mV offset which made a huge difference in temps.

And this way, through the undervolt, it also prevented the cache from getting burned out/degraded (that should receive max 1.35v, theoretically. This opens up a door to a much funnier conversation regarding X3D cpus and how they used to instantly fry when people tried to change the voltage, showing Intel's insane durability of components, but that's a conversation for another time). All 13/14900 builds that I've made in the last 3ish years, basically estimating at least 30-40 builds, are all still functional and non-degraded, while acquaintances and friends that had such cpus and didn't want to bother with undervolt because "manufacturers know better than you, a builder and overclocker"… well , many of them had their cpus replaced.

I strongly suggest you look back and figure out why my reaction was as it was, and maybe, just maybe, if it wasn't based on your own behaviour, comparisons made, and how you formulated things. Overconfidence can be a negative trait, when dealing with someone that understands and knows what you're attempting to talk about. If you need help understanding, I suggest you take a look at the Dunning-Kruger Effect.