Brand New Helios 18 overheating to 105C, frequent BSODs after purchase. Is this normal?

MarkoP
MarkoP Member Posts: 2 New User
edited August 8 in Predator Laptops

Hello, I have a brand new Helios 18 AI (Ultra 9 275HX, 5090) and CPU temperature has been at 105C since day one, whenever doing something gaming related. I've also gotten 5 BSODs in the last 5 days, which I think are related.

I am right to assume this temperature is not normal and I should return the laptop for repair? My old laptop was an ASUS Strix 18 (2023) and the CPU temperature never went above 95C.

Answers

  • Puraw
    Puraw ACE, Member Posts: 18,616 Trailblazer
    https://community.acer.com/en/discussion/727634/brand-new-helios-18-overheating-to-105c-frequent-bsods-after-purchase-is-this-normal

    Hi MarkoP, For your Helios 18 PH18-73 (Ultra 9 275HX, RTX 5090), those 105°C CPU temps and BSODs are definitely abnormal. Here's what you should do to stabilize thermals and improve system reliability:

    🔧 Recommended Fixes

    1. Use the Balanced Power Plan in PredatorSense Avoid Extreme or Turbo modes unless absolutely necessary. Balanced mode helps regulate CPU boost and fan behavior more efficiently.
    2. Update BIOS to Version 1.09 (dated 2025/07/11) This version includes an EC fix that improves thermal and power regulation. You can find it on Acer’s official support page.
    3. Switch to “Best Power Efficiency” in Windows Power & Battery Settings This reduces background power draw and helps cooling, especially during idle or light tasks.
    4. Disable “Processor performance boost mode” in Advanced Power Settings Go to:
      • Control Panel > Power Options > Change Plan Settings > Advanced
      • Under Processor power management, set Processor performance boost mode to Disabled for both On battery and Plugged in (if this option is hidden you have to unhide it by changing the registry key Attributes Value from 1 to 2, see picture below).

    Disabling boost mode prevents aggressive turbo behavior that spikes temperatures unnecessarily. Combined with the EC fix, this should bring temps down significantly and reduce BSODs.

    Let us know how it goes!

    Turbo boost reg.jpg
  • Ugly
    Ugly Member Posts: 34 Die Hard WiFi Icon

    If you've only just purchased this laptop and you live in a country with half decent consumer protection send it back to the retailer for refund or replacement. I've seen a review of this new laptop and it seems Acer have made no effort to introduce underclocking or custom fan curves or a vapour chamber. Just relying on the same fans, liquid metal and heatsinks that were in the PH18-72

  • MarkoP
    MarkoP Member Posts: 2 New User
    edited August 7

    Using Balanced Power Plan makes the temperature much more manageable (usually 82-93C), but sadly it does not resolve the BSODs. They don't happen as often, but they still happen. I've upgraded BIOS to latest, as all the drivers that were available.

    Even if the BSODs did not happen, I would not be happy limiting my laptop to 3.6-3.8Ghz speeds (which seems to be the limit when using Balanced profile). It appears slower than my 2023 ASUS Strix 18 using this profile.

    @Ugly Sadly consumer protection laws do not apply in my case as I purchased the laptop in the name of my company, but I'll see what I can do.

  • Axxo
    Axxo Member, Ally Posts: 2,270

    A brand new Acer Predator Helios 18 experiencing CPU temperatures of 105°C during gaming, along with frequent Blue Screen of Death (BSOD) errors, might be an issue indicating a potential hardware or software problem. The high temperatures and BSODs are not normal and should be addressed.
    Here's a breakdown of the situation and potential solutions:-

    1. Problem Identification:
    High CPU Temperature:
    105°C is significantly above the safe operating temperature for most CPUs, especially during gaming. This can lead to throttling, performance degradation, and even hardware damage.
    Frequent BSODs:
    Blue Screen of Death errors are often related to hardware or driver issues, and in this case, could be triggered by the overheating CPU.
    2. Potential Causes:
    Poor Cooling:
    The most likely culprit is inadequate cooling, possibly due to a poorly applied thermal paste, blocked vents, or a faulty fan.
    Overclocking/Turbo Boost:
    If the CPU is aggressively overclocked or has Turbo Boost enabled, it could be drawing too much power and generating excessive heat.

    Software Conflicts:
    Certain software, especially background processes or those that heavily utilize the CPU, could be contributing to the overheating.
    Driver Issues:
    Outdated or faulty GPU drivers can sometimes lead to overheating and stability problems.
    Hardware Problems:
    There's a possibility of a hardware defect with the CPU, GPU, or cooling system itself.

    3. Troubleshooting Steps:
    Check Cooling:
    Dust Buildup: Ensure the vents and fans are free of dust using compressed air.
    Airflow: Make sure the laptop has adequate airflow and isn't placed on soft surfaces that can block vents.
    Cooling Pad: Consider using a cooling pad to improve airflow and reduce temperatures, especially during gaming.
    Adjust Power Settings:
    Power Efficiency: Switch to "Best Power Efficiency" in Windows Power & Battery settings to limit CPU performance and reduce heat.
    Turbo Boost: Disable Turbo Boost or adjust it in the BIOS or through software like ThrottleStop.
    -----------------------------------------
    If this answers your question and solved your query please "Click on Yes" or "Click on Like" if you find my answer useful.

  • mikeyjames
    mikeyjames Member Posts: 69 Devotee WiFi Icon
    edited August 11
    https://community.acer.com/en/discussion/727634/brand-new-helios-18-overheating-to-105c-frequent-bsods-after-purchase-is-this-normal

    No, it's not normal if you are seeing sustained temps of 105°C regularly and seeing BOD If you are seeing spikes that last a second or so and only happen occaionally, then it quikcly drops down into the 70s or 80s, I wouldn't worry too much as the thermal and power limit is kicking in and doing what it is supposed to do and it may even be just one p-core.

    Whether you purchased as an individual or company the warranty should still apply. Thermal and Power Limit throttling should be stopping your CPU getting to 105°C.

    I have the PH18-73 with the RTX 5080 and running 3DMark Time Spy Extreme Stress Test, with laptop set to Turbo and Max fans. After 15 min this is what I am seeing reported via the RTSS Overlay with HWiNFO 64, plus verifying later with the monitoring min/max recorded.

    CPU

    CPU Package Temp = mostly in 70s and 80s with occasional jump to 102°C for a half a second to max of 1 seond (I ave polling set to 500ms and 90% of the time the temp had dropped substabtially within half a second)

    Core Max = Mostly in mid to high 70s with occasional jump to mid 102°C for half to one second. Checked HWiNFo 64 and it was only on two cores, P-cores 11 & 12, which hit 102°C. The next highest was 88°C.

    Core Distance to TjMax = lowest I saw was 3°C. Checked later on HWiNFO 64 and it was only on the same two cores P-Cores 11 & 12. The closest any other core got was 12°C away from its limit (93°C).

    Judging from this TjMax must be 105°C

    Core Critical Temp = zero cores reported reaching critical temp.

    CPU Package Power Draw = mostly 50-75W, with occasional spikes up around 130W-140W. This barely happened and when it did it was very short lived.

    Power Limit Throttling = Yes on all cores

    Thermal Throttling = Yes on 4/8 P-Cores and 4/16 E-Cores. Interestingly, the same four E-cores (3,4,5,& 6) seem to always thermal throttle, even when the laptop is idling on balanced mode they all regulalrly get to around 80-85°C and constantly throttle. No other E-Cores do this, and I have tried various loads to see if it was just 3DMark hitting these cores and leaving the rest alone. All other E-Cores sit between 8-10°C cooler for some reason.

    P-Cores Low / High / Average

    P-Core 0 = 1695 / 5188 / 4919

    P-Core 1 = 1695 / 5188 / 4984

    P-Core 10 = 1695 / 5188 / 5117

    P-Core 11 = 1695 / 5188 / 3880

    P-Core 12 = 1695 / 5188 / 3979

    P-Core 13 = 1695 / 5188 / 5086

    P-Core 22 = 1795 / 5188 / 4825

    P-Core 23 = 1795 / 5188 / 4960

    I'd love to know what is going on with cores 11 and 12.

    GPU

    GPU load = constantly 100%

    GPU Core = 2200-2400, with occasional spikes to around 2700

    GPU Memory = 1775 pretty consistently

    GPU Power = 173-174W

    GPU Rail Power = Max 180W

    GPU Temp = 74°C max

    GPU Memory Junction Temp = 78°C max

    GPU Performance Limiters = Current only.

    Other System

    System Power Draw = Max 316W

    DIMM1 = Max temp 46°C

    DIMM 2 =Max temp 47°C

    System Drive: Samsung 9100 Pro 2TB (PCIe 5.0) = Max 52°C **

    Games Drive: Samsung 990 Pro (PCIe 4.0 x) = Max 35°C **

    When I had my old Helios 500 you were able to use Intel software to set negative voltage offests. My God it made a difference. Too many people likely did stupid stuff and tried to blame Acer, so they have taken away the ability to do it now, which is very disappointing now. This combined with the debacle of Acer initial press releases claiming the RAM would be 6400Mhz, then finding my laptop only has 5600Mhz RAM installed that only runs at 4400Mhz, means it will likley be my last Acer laptop.

    As I said, if you are just seeing quick spikes occasionally, I wouldn't worry. As you can see from my results, this laptop is sometimes pumping out over 360W, which is a lot of heat to dissipate in a laptop. The throttling should be kicking in quickly to prevent damage. You can see with my results that despite this, 6/8 P-Cores still average up near 5Ghz and the GPU never throttles for any reasdon other than it's 175W power limit.

    I'd download HWiNFO 64 and 3DMark and run the stress test. Monitor for CPU and GPU temps, average CPU speed for your eight P-Cores, temps for GPU core, plus core and memory speeds. If you're seeing lots of thermal throttling with CPU P-cores and low average speeds for your P-cores, and/or GPU throttling for any reason other than power limiting and/or memory or GPU speeds lower then mine, then there must be something wrong, and I'd follow up with Acer.

    Also, look at the report for "Core Critical Temp" and see irf you have any cores reported as reaching this level.

    Maybe also consider where you are. Do you live somewhere where your room temperatures are cracking 30°C or higher, there isn't a high-performance laptop I'm aware of that will do well under these conditions.

    Have you installed any weird software that operates at system level? Maybe uninstall it first. If you have the ability to back up your system drive with something like Acronis or whatever, maybe try something like "Driver Easy". I've never had a problem with it, but wouldn't do it unless I could back up first. Comparable products from Norton and CCleaner have totally screwed my system in the past. Acer is hopeless at keeping drivers updated. Driver Easy found over 20 outdated drivers. There is also Intel's Driver and Support Assistant that has never caused an issue for me.

    With a new gaming laptop in a decently cooled room, you should be able to use "Turbo" settings as long as the fans are high enough, that's what they are built for and cost so much. You shouldn't have to worry about power plans that stress the machine; you paid a lot of money for something that should be able to handle being stressed. While worth checking, unless you live in a rubbish dump there shouldn't be any issues with dust build up, etc with such a new machine.

    Good Luck

    Cheers

    Mick

    ** As far as drive cooling goes, I have tried it all. The best solution by far I have found is https://tinyurl.com/4rjzdevc

    It even kills Samsung's own drives that come with cooler attached by a long way. I always use two together for best results - just use 1 x thermal pad on the drive and two of the copper heatsinks straight togetherr with nothing bwteen them. Use all the bands supplied to be sure.

    My Samsung 9100 Pro was getting up near 96-97°C when being benchtested on a loop with CrystalDisk. After installing the solution, it barely cracks 65°C. The 990Pro (PCIe 4.0) wasn't so much of a problem. It was in the high 70s to low 80s under severe stress but now rarely gets over 60°C.

Categories