980 Ti - vote for my destiny

I'll fire up Arkham Knight for about an hour (1080p GeForce Experience optimized settings), then check Afterburner tomorrow.

Neither is Witcher 3 for SLI since many of the functions like tessellation on their engine only renders on GPU1. It's a great single card stress test though.

DejaWiz said:

I'll fire up Arkham Knight for about an hour (1080p GeForce Experience optimized settings), then check Afterburner tomorrow.

Click to expand...

You will not push the gpu to any where near max at 1080 in that or hardly any other game. In Arkham Knight, gpu usage is only around 80% at most and drops below 70% quite a but when I have tested.

You are going to have to use DSR and test games at 1440 or 4k if you want to push your 980 Ti.

However, for me it has shown instability where other games did not.

Just ran some Firestrike benches:

Core +100
http://www.3dmark.com/3dm/9060844

Core +110
http://www.3dmark.com/3dm/9061063

Core +120
http://www.3dmark.com/3dm/9061027
...1540 peak core clock

something is wrong with your results, lower performance with +110?. what is the maximum reported boost clock?... also a mild OC'd 980TI should be above 21K generally.. at 1400mhz I score above 22K.

on the other hand, running that puppy 3770K at stock? overclock that baby is begging you.. at 4.5ghz my 3770K score a bit over 12K physic score.. consider it..

Araxie said:

something is wrong with your results, lower performance with +110?. what is the maximum reported boost clock?... also a mild OC'd 980TI should be above 21K generally.. at 1400mhz I score above 22K.

on the other hand, running that puppy 3770K at stock? overclock that baby is begging you.. at 4.5ghz my 3770K score a bit over 12K physic score.. consider it..

Click to expand...

I just noticed that and the core clock was reported as 1190 with the +110 test, so I ensured it was +110 and re-ran it. New result posted.

Still unexpected results, are you running any heavy or special thing in the background?, the fact that two of your results present "TIME MEASURING INACCURATE" talk about a issue with the machine, can you run stock clocks and see how it behave? time measuring inaccurate is one of the 3dMarks anti cheat tools, and you have probably a tool messing with the windows timer during the benchmark.. (possible instability maybe?) that's why I ask for a stock run to discard any other problem, have you seen any black blinking during the test? probably driver recovering too fast that it affect the score or the timer.. IDK.. is weird.

You know...I was wondering what the hell that meant. The only things I was running in the background were Afterburner and Asus AI Suite (for monitoring CPU temp and CPU/chassis fan speeds).

I'll have to retest tomorrow if I can. Need to get some sleep in for work tomorrow.