When the CPU or GPU is running at normal temperatures, it is easy to degrade performance, even if it is normal. Most of the time, especially if you are playing games, it will be normal to have some performance drop when the temperatures go above a certain level.
When you start to feel an unusual performance drop or feel that your computer performance is not as good as it used to be, the first thing you should do is check out the temperatures. If they are high, you should immediately do something to lower them.
If you plan to build a gaming PC, you already know how important it is to get the best graphics card for your money. Unfortunately, the best graphics card doesn’t always mean the best performance. If your system is too powerful, then it can throttle itself to lower the temperature. So what is a bad temperature for a GPU?
Why do we even need to worry about the temperature of our GPU? The thing is that it has thermal resistance, which means that there’s some ideal temperature for it. So if you want more performance, you can push its limits towards overheating.
But why would I want to overheat my GPU? For example, mining cryptocurrencies like Ethereum. When your PC mines cryptocurrency (GPU only), the power consumption doesn’t matter much because you’re getting paid for it anyway.
If you increase performance, then decrease power consumption somehow (overclocking or undervolting), you’ll be able to mine more with less cost! Of course, this comes with increased heat, which might result in damaging your card if not managed properly (or lead to decreased performance if the power limit is reached).
And if you think about it, this is exactly what happened with all those ETH mining rigs that ended up becoming fire hazards because of poor thermal management!
If you want better performance, there’s a certain temperature for your GPU. If you push it further, its performance will decrease until the power limit is reached or something breaks. So be careful with overclocking or undervolting unless you’re experienced.
Remember that GPUs are pretty much just as good as your CPU cooling capabilities may allow them to be! So either get better air-cooling (e.g., an AIO), buy a better CPU cooler (e.g., Noctua NH-D15), or get a water-cooling setup (e.g., go with EKWB).
Another thing to remember is that overclocking voids your warranty. If you damage something while overclocking, you won’t return it because it was working on factory settings before.
If you want better performance from your GPU, then I suggest getting a more powerful one instead of overclocking a less powerful one, unless you know what you’re doing and have low expectations for some reason. Most GPUs can reach 80°C+, but some don’t even break 50°C! Be careful with the “silent BIOS,” which sometimes doesn’t allow changing temperature limits.
Finally, remember that overclocking or undervolting will only affect the power consumption and the performance of your crypto mining rig if you use it for cryptocurrency mining. So if you don’t want to mine, don’t touch those settings!
A bad temperature for a GPU is very subjective, as there are many factors that could alter the temperature of a GPU. If you are using a custom cooling solution or a closed case, there are very few times where a high temperature would be considered ok.
The high temperature may not even be caused by the GPU but could be caused by other components in your computer. The ideal temperature for a GPU is between 50 to 70 degrees celsius. If the temperature is above that, it is time to invest in a cooling solution or open your case.