Originally Posted by IMKR
Originally Posted by mcg75
The amount of stress Furmark puts on the card is unrealistic and far more than any game or benchmark will.
but what exactly is the difference between
doing a stress test on a GPU (furmark)
and benchmarking a GPU? (3dmark)
it seems like its the same thing, (stress test and benchmarking)
They can be used for the same purpose, yes.
3Dmark is more realistic to a gamers general and daily load, maybe a little harder.
Furmark used to be a good method for testing GPU's but there was a short falling out as the application began to fail to produce proper results when other forms of stress would do the opposite of furmark.
For example: I have a 290x.
If I furmark it will artifact and crash around 1070MHz
If I 3Dmark I start for crashing to desktop around 1170. - There is NOTHING ELSE other than furmark that causes me to have issues with this card like this.
this could also be the other way around, and also true with things like folding, mining etc.
It just doesn't happen to be a viable testing solution to some anymore.