As I get deeper and deeper into blender I’m beginning to have some concerns about the wear it has on my computer.
Leaving my computer running for twenty minutes at full processing power is giving me some doubts. By regularly using blender and rendering stuff, how much will that shorten my computer in general? Will it have specific effects on my graphics card?
Any info on this would be helpful.
so long as your box stays sufficiently cool, lifetime should be unaffected
In fact, turning the computer off and on is worse than just leaving it running
Somehow rendering all 4 of my cores at maximum for 25 minutes doesn’t seem like it would stay cool.
I wasn’t really looking for computer advice (I almost never turn this computer off btw) but more advice on how blender would affect it.
That’s not really separable, is it?
However, rendering produces heat, either in the CPU or the GPU according to your render device of choice. That’s not a problem as long as you provide a sufficient cooling solution, as heat is in fact the one thing that will significantly reduce the life cycle of your components (spontaneous defects from manufacturing errors aside).
So, get yourself a program to monitor CPU/GPU temperatures (CoreTemp, SpeedFan, etc.) and render on. If the temperatures stay below the critical values for your respective components/models you should be on the safe side. If not, it’s time to review your PC cooling solution - just as daren already suggested.
You are worrying here if your machine will die from thermal stress in 100 or in just 75 years really.
Sure there’s always the chance that a part will die to heat, but a computer is a tool. Tools wear off if you use them.
And they are one of the few things still built to last, simply because they got a short life cycle due to a lack of computational power after a relatively short period of time, creating returning customers. If the power would be enough for a decade I am sure CPUs/GPUs would have planned obsolescence…
As long as the T-max of your components stay below the manufacturer specifications it should be no problem to run 24/7.
My machines often run for weeks without a pause with 100% load and there’re no issues besides a room with warm, dry air.
Continuous thermal stress is less of an issue than repeating thermal stress in short bursts - obviously the expansion and retraction from heating cooling can cause micro fractures. But that’s really more of an issue with Notebooks and other mobile devices. In my past 25 years I haven’t really encountered any case of a regular graybox failing due to thermal stress.
The one rule to obey: Make sure you got a good airflow in your case and expel unwanted heat.
Every cooling system can only cool down a delta temperature relative to the surrounding air.
Easy example: Component 70°C, Air in case 50°C, Delta power of cooler 40°C
Obviously you could cool your component to 70-40=30°C.
It’s impossible though because you can’t cool below the temperature the surrounding air has. 70-x=50 -> x=20°C basically leaving you with wasting half the possible cooling power - and to make matters worse, the air inside the case will continue to heat up, especially if you got a decent metal case and it starts to heat up as well.
Also keep in mind when cooling the case that you can either have a high pressure or low pressure air cooling.
For low pressure you have more air volume expelled than shoveled in.
For high pressure you have more air volume intake than you expel.
Relevant? Yes. High pressure is better. If your fans (with dust filters) shove in more air than you expel you get a high pressure system in the machine and all the gaps and holes in your case blow air keeping dust out of the machine.
If you expel more air than you intake all those gaps start to suck in dust.
And dust is a very good thermal isolation preventing proper heat transmission for cooling.
Modern fans are labeled in m³/h so it’s easy to find out which system you got.
hth
If you’re worried about your computer that much then you shouldn’t use it! Turn it off, put it in a nice vacuum-sealed container and store in a dry, cool place. Never, ever turn it on. 
That’s a question I’ve been running into. How exactly do I tell if I’m using a CPU or a GPU? What’s the difference between the two?
My Coretemp box is showing my CPU usage, so I’d assume I’m using CPU. Don’t most computers come with both or is it more of a “Do you want a CPU or a GPU” when you go to buy one?
Well, every PC has a CPU (= central processing unit). The GPU (=graphics processing unit) is a dedicated secondary processor used in graphics hardware (graphics cards and gaming consoles). A hybrid of those two is the so called APU (=advanced processing unit), which is a graphical co-processor directly integrated into some CPUs.
So, if your PC contains a dedicated graphics card, you for sure have also a GPU in your system. If you’re rendering with Cycles, you can choose between CPU and GPU rendering, as long as your GPU is made by nVidia (ATI/AMD cards don’t work). Depending on your GPU type and model GPU rendering can be significantly faster than CPU rendering.
Blender defaults to CPU rendering, so - if you did not deliberately switch to GPU rendering - you’re indeed using CPU rendering for now.
Hi, it is possible to render with Cycles on the GPU.
This is working on Nvidia cards only and make sense if you have a good card > 150$.
Heat is bad for all electronic parts, keep your CPU < ~65°C and the GPU < ~80°C and you don´t have problems.
Cheers, mib.
EDIT: Ups, IkariShinji was faster. 
Where would I find this option in blender?
In the Render panel - but it is only an option if a supporting GPU is detected in the system.
Hi Meed96, it is may time to post your system specs.
Cheers, mib.
Switch to Cycles as the default renderer.
Under User Preferences -> System set Compute Device to CUDA and select your graphics card. If you can’t select CUDA or no graphics card is shown, your device is not compatible with Blender’s GPU rendering.
Under Render -> Device switch to ‘GPU Compute’
Thanks, I’ll do some testing.
