Extended high temps and risk of physical damage

Finished my first PC build recently and have been going thru various benchmarks, When I run Geeks3D GPU tester benchmark, my slightly OC'd Nvidia TitanX Pascal hits 89 Celsius and my fan ramps up to maximum.
This is only for a few seconds but I am left wondering what possible physical damage may occur to either GPU or CPU when exposed to high temps for prolonged periods, from several minutes to several hours.
Anyone?

89c won't hurt anything in the short run but it may be a little high for hours at a time on a frequent basis I'm thinking. Tell us about your computer case and the ventilation characteristics. Make and model of the case? How many intake and exhaust fans do you have and where are they located? How's your cable and wire management? Hot air being trapped inside the case will contribute to hotter gpu temps.

Looking at this chart from a review of this card and it looks like max temps under load were not much over 60c: http://www.bit-tech.net/hardware/graphics/2016/09/20/nvidia-titan-x-pascal-review/11

A picture is worth a thousand words so if you could attach (not link) pics of the inside of the case it would be helpful. Use the "Insert Image" tool at the top of a new post window to do this. Pics must be under 1 mb in size I think to be attached so if they need downsizing or resampling, use something like Irfanview to reduce the pixel count.

Attached is a screen grab from the official ekwb install video. My build is almost exactly the same. The only 'clearance issue' I had is both cpu power cables are just brushing up against the side of the radiator (obscured in the photo)

A canned photo like that does not show me anything about YOUR fan placement and cable management. If were to take that photo as indicative of what your system's ventilation is like with the sides and front panel installed I could tell you immediately what the problems is since there are no ventilation fans installed in that case.

You mention slightly oc'd, did you change any voltage settings at all? A picture would be great as stated above.

Return it to stock as well and see what your temps are like.

You could run it 90c 24/7 for the 3 year warranty it should make, it also it will clock down when the GPU is in danger. People that do distributed computing run them 24/7 80-90c till they sell them to someone else.

trents said:

Looking at this chart from a review of this card and it looks like max temps under load were not much over 60c: http://www.bit-tech.net/hardware/graphics/2016/09/20/nvidia-titan-x-pascal-review/11

Click to expand...

Those temps reflect the Delta T....as we don't know what their ambient temps are it would likely be safe to add in the area of 20 or so degrees to that

Ok so the only real benefit to watercooling then, is NOT to prolong the life of your components, but rather to prevent throttling that happens at high temps?

Have you tried a custom fan curve?

I set mine really aggressive...100% at 80 C. The card usually tops out at 81 C with a 1925 GPU clock.

eco_bach said:

Ok so the only real benefit to watercooling then, is NOT to prolong the life of your components, but rather to prevent throttling that happens at high temps?

Click to expand...

Yes water cooling is for Turbo (overclocking) not longevity.

If you don't like the fan profile then water cooling is better so it won't down clock from turbo modes.

Turbo is based on temperature and voltage.

Graphics card temperatures typically range from 30c to 40c at idle and from 60c to 85c under load. Most high end video cards typically have a maximum temperature between 95c 105C, at which point the system will shut off to prevent damage. A video card is not considered to be overheating until it exceeds 90c under load. At this point, artifacting may start to occur up to the point of thermal shutdown. http://www.evga.com/support/knowledgebase/

Click to expand...