Predator helios 300 G3-571's 1060GTX being throttled at 65 degrees Celsius by performance limit

DatBoi
DatBoi Member Posts: 1 New User
edited November 2023 in 2019 Archives
Hello,
I recently bought the G3-571 laptop. The GPU seems to have a temperature limit of 65 degrees Celsius which is absurdly as it should be 85 degrees celsius. I have updated the bios to the most recent version (1.21) and it still suffers MAJOR throttling at 65 degrees. It appears that multiple performance limits are going off for the GPU as well. including Power, Reliability volatage, and utilization. The GPU usage drops to about 50% to keep temperatures near the limit. I have lowered the CPU voltages by 95.7mV to keep the CPU at a lower temperature but nothing seems to work. The file I attached is a folder of three screenshots showing the graphs of the temperature and usage of the gpu as well as the specific performance limits that are being triggered. I cannot Find a work around for this.
 

Answers

  • JackE
    JackE ACE Posts: 44,848 Trailblazer
    The battery must be near full charge and plugged in. Control Panel power plan must be set to high or max performance. Don't under or overvolt. Jack E/NJ

    Jack E/NJ

  • Bixtape
    Bixtape Member Posts: 4 New User
    I had absolutely the same problem bro. I bought it a week ago ever since I got it I'm experiencing fps drop on black ops 4 with the same sensor report in hwinfo. I tried absolutely everything , been discussing with the helios 300 fb group for 2 days , trying to fix or atleast find exactly what is causing it with no success. I tried with undervolt , stock voltage , different nvidia drivers , locking the frequency of the gpu to lower mhz than stock ,because in stock it was hitting almost 2000 mhz which is not supposed to do so without an overclock. I believe it is a gpu malfunction , where the gpu works at really high voltage and frequency and the system is not able to keep it in a stable state therefore the fps drop occurs! And yes I was playing on 100% battery and it was dipping to 90% after some time of gaming ,, whatever this means. I'm getting a replacement and will see what happens and will let you know the end results!
  • tobimaru
    tobimaru Member Posts: 315 Skilled Practitioner WiFi Icon
    Pascal designed cards are designed to step down their core clock speeds at anything above 60C. From 60-85C you should be seeing gradual loss of core clock speed in 13MHz increments as GPU Boost 3.0 works. This is normal operation.

    Within GPU-Z, when you see the 'PerfLimit' reason being "VRel, Power, or Utilization" these are not errors. There is always something that is going to limit the boost ability of the card.

    When you game or run a benchmark what are your peak CPU & GPU temperatures? At what temperature does the GPU utilization drop? Typically the GPU being underutilized is a sign of the CPU being overworked or otherwise bottlenecking the graphics card.
  • Bixtape
    Bixtape Member Posts: 4 New User
    tobimaru said:
    Pascal designed cards are designed to step down their core clock speeds at anything above 60C. From 60-85C you should be seeing gradual loss of core clock speed in 13MHz increments as GPU Boost 3.0 works. This is normal operation.

    Within GPU-Z, when you see the 'PerfLimit' reason being "VRel, Power, or Utilization" these are not errors. There is always something that is going to limit the boost ability of the card.

    When you game or run a benchmark what are your peak CPU & GPU temperatures? At what temperature does the GPU utilization drop? Typically the GPU being underutilized is a sign of the CPU being overworked or otherwise bottlenecking the graphics card.
    So you are telling me this is normal? 
    See attached photo.
  • Bixtape
    Bixtape Member Posts: 4 New User
    @tobimaru Just fyi , I have the same specs , Gpu timing in game jumps to 150 and CPU timing stays below 15 , do you still think it's a cpu bottleneck with temps of 80C CPU and 74C GPU
  • tobimaru
    tobimaru Member Posts: 315 Skilled Practitioner WiFi Icon
    edited December 2018
    At those temperatures (80 CPU/74 GPU), you will definitely be within the operating range of GPU Boost 3.0 limiting your clocks speeds. Even at 65C there will be a small downclock compared to maximum possible boost clocks. It is also not uncommon for you to see sporadic clocks ABOVE your max boost limit IF your thermals are low enough. All that said, these cards are very dynamic and there is not exactly a way to "lock" a clock speed for every scenario.

    Have you configured nVidia Control Panel power settings? Windows 10 Power Plan? Predatorsense Fan profile?

    Enabling VSYNC and other taxing nVidia features is an easy way to limit the GPU's utilization. I will see a 30-40% utilization drop in GPU when enabling certain features within nVidia Control Panel. When vsync and other features are disabled, the GPU is back to 100% utilization in-game. Spend some time in the nVidia menu and see if it helps your performance.

    EDIT - It's impossible to diagnose an entire system from one in-game screen shot. What I can see is your CPU is only running at 2.2GHz. The card is severely down-clocked, and the temps are fairly normal. All of this indicates a low power state such as being unplugged. There are several possible answers; the laptop is unplugged, the power state is configured incorrectly within Windows, the AC charging system and/or AC adapter is faulty.
  • Bixtape
    Bixtape Member Posts: 4 New User
    That's a screenshot when the fps drop occurs , that's not my normal stats . I've tried so many things ,including everything you mentioned. I highly doubt that running stock settings will experience things like this , that's not normal. How can u explain the high frequency that the gpu reaches without overclocking? ( almost 2000 mhz) it has happened even when the gpu temp is at 69C , no difference of how long the fps drops or how many frames are dropping. I really can't agree that running stock settings with lowest graphics in game should experience in this. There's no place to think I'm playing without the laptop plugged in the power outlet , I'm not such a newb.
  • DGB
    DGB Member Posts: 1 New User
    Bixtape said:
    That's a screenshot when the fps drop occurs , that's not my normal stats . I've tried so many things ,including everything you mentioned. I highly doubt that running stock settings will experience things like this , that's not normal. How can u explain the high frequency that the gpu reaches without overclocking? ( almost 2000 mhz) it has happened even when the gpu temp is at 69C , no difference of how long the fps drops or how many frames are dropping. I really can't agree that running stock settings with lowest graphics in game should experience in this. There's no place to think I'm playing without the laptop plugged in the power outlet , I'm not such a newb.
    A little late on this one, but here's my story. TLDR at the end lol.

    I bought The Helios 300 571 in Jan of 18. It ran like absolute garbage out of the box. The first issue was, you guessed it, the heat. First thing I did was play SWBF2 and withing ~15 minutes it was burning the hell out of my hands, and throttling, and actually hit 99+ and crashed. I was out of the computer game for awhile, and never have been a laptop guy, so I didn't know what to do. My friend said CPU undervolt, so I did that with Throttlestop and it thankfully brought my temps into non-hell levels. It still got hot, and I have a cooling pad/max fans on etc. 

    I should have probably sent it in, but I was able to get it to work and I needed it, and throttlestop kept it from getting too hot ~85 celcius max. I've used the laptop like this to do work and game for over a year. It jutters and microstutters/not so micro stutters, and I knew a lot of other people had this issue, so I honestly just chalked it up to being a piece of *****. 

    So that brings me to now. The laptop was still playable, and pretty good overall. I recently was playing DS3 and everything looked ok I guess, the small stutter/lag didn't effect game play enough to matter. This week I picked up an Oculus Rift and I knew something was wrong. It was almost unplayable. Through trial and tribulation I was able to get it working more smoothly, but realized the rift wasn't the main issue, it was my laptop. Optimized the rift as well, but that's another story. 

    So, I spent all night doing a bunch of *****. Set up Throttlestop with a better profile, Overclocked with afterburner(turned off monitoring), reinstalled VGA, IO, and SATA drivers, updated the bios (1.13 -> 1.21 lol). I'm sure all this helped, but NOTHING like this video I'm about to link.

    https://www.youtube.com/watch?v=XTPMwQKlfPE

    Follow this ***** video. Maybe one of y'all smart guys can learn me something, but all I know is I followed and it lead me to water. 

    Optimizing this laptop for gaming is an absolute must. I turned off what felt like every windows setting, but one in particular stuck out. The gaming settings. There's no way Win10 knows how to optimize for gaming lol. Turn that trash off. Anyway, after going through the video, I skipped the optional steps, I booted up DS3 and it was like I had an entirely different computer. It was smooth with NO stuttering. Max graphics rock solid 60 fps. Even in big battles there was barely a spike, insanity. Not to mention my temps are well within control now. It just doesn't get as hot, normally hanging around ~75 Celsius even while gaming.    

    The Rift also was heavily improved, everything was. I do think this laptop should be a lot better out of the box, but I'm very glad it can still be salvaged. I reran UserPCBenchmark and the cpu is running in the 99th percentile now. Unbelievable. Somehow, even the textures and menus look better and clearer. Anyway, follow that ***** video, and don't be a lazy ***** like me.

    TLDR: Heat Problems? Laptop performance blows? Watch this video: https://www.youtube.com/watch?v=XTPMwQKlfPE 
    It helped me!