How to lower your GPU temperature in 2024

How to lower your GPU temperature in 2024
Meghan Coon Updated on by

Video Gamer is reader-supported. When you buy through links on our site, we may earn an affiliate commission. Prices subject to change. Learn more

Knowing how to lower GPU temperature can be instrumental in solving performance issues you may be experiencing while gaming on your PC, and it can offer a relatively quick fix before you commit to buying new hardware you may not need. 

An overheating GPU manifests itself in thermal throttling, which is essentially when your graphics card exceeds its maximum temperature, which CG Director explains as around 90-95 degrees Celsius, and begins to slow itself down to reduce heat being created. You’ll notice when this happens as it tends to lead to random shut downs, quality issues, buffering and blue screening, which is not brilliant when you’re in the middle of a crucial campaign. It can even lead to long term damage of your hardware if not treated.

Why your GPU is overheating

Learning how to lower your GPU temperature is one thing, but it is also necessary to identify what it is that could be causing your GPU’s temperature to rise.  This could be down to a lot of factors that may not be a direct hardware issue; the general upkeep of your computer and its surroundings also play a role in affecting the heat of your GPU. 

 For instance, the external temperature of your PC, especially if you live in a hot climate or if you notice your GPU temperature rising most during summer months, will affect the efficiency with which your GPU distributes heat, as it will have to battle both the outside heat and that being produced internally. If your PC is located in a warm environment, it’s important to ensure that the space surrounding the computer is not cluttered and the fans are not being blocked. 

As well, if your PC is not being cleaned regularly, this will lead to a buildup of dust that can prevent easy airflow throughout the PC, which will ultimately reduce cooling and will contribute to hardware overheating. XDA Developers points out that disorganised cables and/or the positioning of your GPU could also be causing obstructions to the airflow around your PCIe, which prevents heat from being effectively filtered.

Though everything you use your PC for will use your GPU to some extent, even watching high definition videos will not raise your GPU higher than 80 degrees Celsius, even if you are not running on a powerful GPU; however sustained, demanding workloads such as rendering or running GPU-intensive games for extended periods of time (longer than just a couple of hours) this will put strain on your GPU and could lead to overheating and throttling. To keep a better eye on your GPU’s temperature while you’re gaming, Windows 11 allows you to monitor and optimize your GPU from Task Manager. So how do you lower your GPU temperature to avoid this? Well, there are a few ways I would recommend. 

Stop overclocking your GPU

Overclocking is described by AVG as gradually increasing the clock speed of your video card to boost its performance, increasing its core frequency to raise its processing power. This, by definition, is increasing the power consumption of your GPU which leads to more heat being created. CG Director points out that if your overclock is stable regardless of running at a temperature higher than around the 80 degrees Celsius mark then that’s OK, but it is worth checking to ensure that overclocking isn’t the root cause of the overheating.

Undervolting your GPU

Undervolting is when you decrease the voltage supply to your GPU while ensuring stability in the stock clocks, a bit like overclocking but the other way. This can be achieved by using the manufacture-provided tools or a third-party software like MSI Afterburner. This should be done gradually and XDA Developers recommend this process should be followed by thorough stability testing to reduce the risk of crashing. This is a bit of a slower process to reduce the overall temperature of your GPU, reducing the voltage supplied to your graphics core will help keep the heat produced by your GPU a bit lower. 

Change your GPU’s thermal paste

Thermal paste is a core component when adding or maintaining a GPU cooler, even though the process can be a bit of a pain. If you’ve ever built a PC or modified your hardware you have probably experienced thermal paste before, and while GPU heatsinks come with it pre-applied, this can sometimes be of low quality or wear away over time if overheating is an issue for your PC, it’s probably a safe bet to remove it and reapply. Needing to change your thermal paste could be down to a number of factors, as I just mentioned, lower quality thermal paste (as is typically the case with lower quality things) can decay faster and conduct heat less effectively than others, this means that it will require changing more often to prevent your GPU from overheating. 


As well, if the paste was applied manually, there’s a chance there may be air bubbles in it or if it was poorly distributed, then it won’t be transferring heat as effectively as it would had it been applied correctly. Thermal paste can last several years depending on you use your PC, smaller workloads won’t raise your GPU temperature too high or too often, but if you are an avid player of AAA title games or if you spend long lengths of time rendering big projects, then the efficiency and durability of your thermal paste won’t last more than a year or two, so changing it can be detrimental to ensure the highest possible efficiency for heat transfer to your cooler. If you have never applied thermal paste before, check out my article on how to apply thermal paste for a step-by-step guide.

Lower your in-game settings

While this one may not suit everyone, it is arguably one of the simpler methods to lower your GPU temperature. By reducing the in-game graphics settings, such as the resolution, texture quality or shadow details, you can reduce the demand on your GPU, which means it is not working as hard and therefore reducing less heat overall. This will however affect the quality of your gameplay experience, and may not help a huge amount in the long run if the high temperature is being caused by hardware factors, such as thermal paste or poor airflow within your PC. Simply put, the higher the VRAM cost in demanding software, the harder your graphics card is going to need to work, so be smart with your settings if you’re noticing alarming figures in-game.

Adjust your fan speed

An effective way to combat your GPU temperature getting too high is to adjust your fan settings. If you have never changed your fan speed then it will likely be set to automatic, so that as your GPU temperature rises so should your fan speed. But you can alter these settings in your BIOS or via a third-party software like MSI Afterburner, where you can go into ‘predefined fan speed curve’ and set it to custom. This allows you to set your fan to kick in at a lower temperature (like when your GPU’s temperature goes above 75 degrees Celsius) to prevent overheating. You can also reduce the overall fan speed, as a GPU fan running on 100% can not only increase the risk of the fan wearing-out prematurely, or burning out the motor, it can also add to the heat being produced. Make Use Of found that it could maintain a temperature of around 70 degrees Celsius when the fan was running at only 90% fan speed. So by altering your fan settings, you can keep GPU temperature down while increasing the lifespan of your fan and preventing further risk of thermal throttling.

What GPU temperature is too high?

It is recommended that your GPU should be around 60-80 degrees Celsius, depending on what GPU you are using and how you are using it. If your GPU exceeds 95 degrees Celsius, this is considered too high and when you are at risk of thermal throttling.

Why is my GPU running so hot?

This can be down to several different things, the overall cleanliness of your PC, the positioning of your graphics card within the case, your fans maintenance or it could be because of the condition of your thermal paste, and your cooler, or the age and condition of your actual GPU.

(Image by NVIDIA, image of NVIDIA RTX 40 series)