bench marking...

[kinkywink]

does anyone know if linus/slick actually play all the bench marks they take?

 

i was just thinking it would be more accurate if they designed a macro as such, to run through exactly the samething.. this would eliminate human error and make the results more accurate not saying this is easy to do but yeah

[AlwaysFSX]

Slick does it but he's so precise it's like he IS a macro

[kinkywink]

Slick does it but he's so precise it's like he IS a macro

:L every video they do after a bench mark, for example when the r9 290(x) and 780ti came out they looked like they hadnt slept in days :L just thought they would have found an easier way to do it by now

[AlwaysFSX]

:L every video they do after a bench mark, for example when the r9 290(x) and 780ti came out they looked like they hadnt slept in days :L just thought they would have found an easier way to do it by now

Real world benchmarks require somebody to play it... I think the fifth person they hire should be the one just to do the benchmarking

[GBBnero]

Real world benchmarks require somebody to play it... I think the fifth person they hire should be the one just to do the benchmarking

 

Where can I sign up?

 

Slick actually plays the game to the point where it looks exactly the same as the other benchmark videos. He mentioned that he had trouble bench marking low end cards because he would get teleported or certain events, like explosions, would happen at a different time. 

[AlwaysFSX]

Where can I sign up?

Well Nvidia has launched their G-sync module (I think, on their website) that you can buy for $200

[kinkywink]

Real world benchmarks require somebody to play it... I think the fifth person they hire should be the one just to do the benchmarking

real world bench marks mean non synthetic. real world doesnt require somebody to play it, its just a real life scenario, for example gaming. its going to happen its not just made up to stress the cpu or gpu -

and it wouldnt make a difference if slick is purposely trying to make it exactly the same because that would defeat your proposition, because your practically saying its not a real world example if its identical, but yeah they should hire someone just for the bench marking lol

[IVIeLt]

Pre set benches are never a good indicator of a cards performance.

 

I've had a few cards that would have anywhere up to 50 percent performance hits in real game scenarios compared to the games built in benchmarks. Tomb Raiders benchmark is a particularly bad example of this.

[kinkywink]

Pre set benches are never a good indicator of a cards performance.

 

I've had a few cards that would have anywhere up to 50 percent performance hits in real game scenarios compared to the games built in benchmarks. Tomb Raiders benchmark is a particularly bad example of this.

i dont know what tomb raiders like but im more or less meaning making the bench mark yourself, which would be better then built in benchmarks