7. If your laptop crashes at -100mV, you need to try with -80mV. Keep on increasing the value by 20mV until you are stable without any crashes. Crashing is a part of the process. The laptop will restart itself. This is nothing to be afraid of.
8. If, on the other hand, your laptop is stable at -100mV, start adding -10mV until you start experiencing system crashes. The last stable setup before you experience any crash will be your undervolt limit. My laptop’s limit was -150mV. Yours will be a different value.
9. Finally, click “Apply” and “OK”
10. You can run a benchmark test with “TS Bench”. Take a look at the temperature values. The undervolted temperature values should be lesser than that of the original vales we noted earlier. Also, if you have any error, simply increase the undervolt value by 10mV and run the benchmark again (If I get any errors, I would change it from -150mV to -140mV and re-test it until I have no errors). Here are my temperature values before and after undervolting.
Thank you very much. You have said everything precisely and to the point. Loved it
You’re Welcome! Glad you loved it!
man. this helped me a lot. Thanks for Point to point explanation.
My pleasure 🙂
there are no temperature changes after i undervolted my cpu even up to -165mV. max temp are always 96-98 celcius
Are you sure? There must be at least some kind of temperature difference.
Fine. If it doesn’t work you need to try replacing the thermal paste or getting a better cooler.
So my clock is 1920, and voltage 1038. I’ve dragged everything from 938 upwards to 1920 and not noticed any difference with temperatures (still climbs to 88 degrees).
I tried upping the preceding node (937mV) to 1920MHz, but when I hit apply it jumps all the nodes I just spent ages moving to 1935MHz. Moving them all back to 1920 and applying just did the same thing again. Any idea why it’s doing this / how to stop it?
I suggest reinstalling the MSI afterburner software if the nodes are moving back to their default places. You should definitely notice a difference in temperatures once you adjust the nodes and lower your voltage. If undervolting does not work, changing your thermal paste is the only way to combat rising temperatures. Thanks for reaching out! 🙂
Thx for the guide. Also can i undervolt gpu and overclock gpu at the sametime?
I would suggest you not to do both at the same time. Either undervolt or overclock. We have in-depth tutorials for both.
Hello,
Thank you for this detailed explanation. I have a question. When I kept incrementing, I reached the BSOD for more than -150.4 mV. I wasn’t able to restart my computer until I rested my BIOS. After doing that, I logged in perfectly and checked whether this turned on turbo boost again (which I disabled few months ago). It was disabled, and according to Throttestop, after running TS bench, the temperatures did not exceed 62-63 while they were reaching mid-high 70’s. Note that the disable turbo was unchecked.
I did the same thing today, but even with the previously saved voltages, I had temperatures reaching high 80’s. When I checked “disable turbo” temps got back to the values of yesterday’s test (62-63). it was as if today turbo boost was enabled on its own. But I checked it, it was turned off. Why is this happening? Why do I have today to check “disable boost” while I didn’t have to yesterday?
Thank you so much for the tutorial, my temperatures are not exceeding 75 for both cpu and gpu.
Honestly Daniel, I cannot give you an exact reason until I do it myself. My setup did not have any such issues. I wish I could help you real time but we know that is not possible.
I mostly think it was due to some software bug. Try installing previous versions of Throttlestop and see if this happens again.
Im really happy my tutorial was useful to you. Good luck and stay safe 🙂
Tbh I’m satisfied with the results, I’m gonna keep that box checked and enjoy low temps. I have an msi GS65 stealth, and temperatures are only reaching max 75 for gpu and 69 for cpu. Before that it used to reach 90s for both. I also suspect excess heat have been transferred to my battery and caused it to swell. After replacing it with a new one, I wanted to prevent this from happening again.
Thank you again!
Cheers
Hi vishwak
Will your tutorial work for a suspicious not powered enough PSU ?
I have an evga 500w 80+bronze , never had any issues but recently I bought a brand new rx580 ROG strix gpu.
It is stated that the minimum psu for that gpu is a 500w but I’m facing random freezing and monitor shuts down. But pc stays on. Not even gaming just normal browsing freezes up the screen and shuts the monitor down.
I was thinking undervolting the gpu with your tutorial might resolve the issue thinking maybe the psu is not enough for the rx580 . I’m on a budget so right now I can’t afford buying another psu . What are your thoughts ?
Hey Plaxi,
This tutorial will work on your GPU. You can safely undervolt your RX 580 without any issues. But I doubt if this will solve your monitor’s problem.
You see, undervolting it is useful for improving temperatures and performance. Though it consumes less power than before, I don’t think that it will decrease your power consumption to such a level where you can use a 500w PSU for a 550W GPU. But still I suggest you try it out and see for yourself.
Cheeers!
Hello,
Cant seem to be able to download the HWinfo64 software, it keeps asking me to enter a businessemail (i tried entering my main gmail, but that doesnt seem to be good enough)
Is there another way to find gpu max frequency or can i do the changes in afterburner without checking it first?
Really appreciate any help, i’ve already changed thermal paste but still my laptop (predator helios 300) goes to 95 celcius when gaming.
Hey there!
You can use software like CPU-Z and GPU-Z as an alternative to HWinfo64.
Also, you don’t need to enter any email. Just go to : https://www.hwinfo.com/download/
You will see a green colour “free download” button that turns blue once you click it. There will be few options from which you can download the Software.
Helios 300 is a good gaming laptop but heats up just as much. Try cleaning the dust as well in addition to applying the thermal paste.
Good luck!
will ThrottleStop affect performance and/or ram usage?
No. There won’t be any performance issue. You can use it minimized.
Mate thanks a log for the tip, now i don’t have to buy a new laptop because with a little undervolting it improved a log, thanks
Glad it was helpful 🙂
Well, the program for cpu does not work on amd cpu. Second what about AMD cards?
Hey there!
Throttlestop should work for both AMD and Intel CPUs.
If it didn’t work for you, try Ryzen Master software.
Amazing!!!!!! Great guide!!!! I just learned about thermal throttling this afternoon as my laptop was lagging with Cyberpunk 2077. By following this guide step-by-step I reduced CPU temps by 12°C and GPU temps 10°C, the FPS do not drop anymore!!!
Big Thanks!!!! 🙂
You’re welcome! Happy Gaming!
Years and years of hearing about undervolting CPU/GPU, and most guides were overly complicated. I have had throttling issues with my Intel and Nvidia based laptop for years but never thought I’d ever undervolt the CPU. With this guide, I was able to undervolt my CPU and that pretty much solved my throttling issues. As for the GPU, the Nvidia GTX 1050 voltage was locked but I was still able to edit the GPU clock which I had limited to the max GPU clock (after checking using HWiNFO64). Not sure if the CPU undervolt and/or limiting the GPU clock helped eliminate the throttling… but the issues are gone. My games no longer have sudden frame drops every 10 seconds or after 30 minutes of gameplay, the frame rates tanking considerably. This guide helped tremendously!
I’m glad you found this guide useful.
Happy new year 2021!
I am very excited to give this a go. I’ve spent some time gathering the components to max out performance at my budget level, but now I’m having throttling issues due to overheating.
My setup is this:
X570 Taichi
Ryzen 5 3600 (Noctua Fan)
2x GTX 1070 (SLI installed and used for some games)
The CPU stays below 80 with my Noctua replacement fan
The GPU’s get hot enough to throttle themselves around 82 C (yikes)
One GPU will still get hot if not run in SLI
My question is about undervolting GPUs in SLI. You say that even if the cards are the same, they need different undervolts, but If they are in SLI does that mean that they should have matched undervolting, like a midpoint between their boost volt and undervolt? Is this even possible, or recommended with 2 gpus in SLI?
Hey Joseph!
I’ve personally never tried Undervolting my GPUs in SLI.
I encourage you to try it out and see for yourself. I’m sure that when you use two GPUs in an SLI, they should function as one.
When those GPUs are in an SLI, there is no hard and fast rule that their undervolting values should match.
This is what I would suggest, Try Undervolting one GPU and see if you could reduce its temperature. Next Try Undervolting both in SLI. See what works for you and proceed accordingly.
Good Luck!
Hi i have a AMD threatripper 2950x. ThrottleStop is not working. Is these a displacement software?
Hey Jan!
Try reducing the offset via the BIOS settings to reduce voltage.
Hello! Very detailed for this 74 year old new gamer [1 year] using a wonderful ASUS M15 / RTX 2070 Max Q. I use a great cooler, the Gt300 [Amazon] which has turbines, not fans. I am hitting 78, at short times 80. I use MSI overlay. I am totally computer illeterate but I think, I know I can follow your very clear steps. Should I really undervolt GPU and CPU or just leave it alone! 240Hz 3ms screen and run most of the time 120 FPS +/- and very happy! I also keep my dedicated game room temperature controlled very cool. What do you think? Honestly a bit nervous doing it! But I hear 80 max is okay. I appreciate.
Hello Ara!
I’m happy you found my guide useful.
Yes. Max 80 for a laptop is okay. Still, I would suggest undervolting for better performance in the long term.