arestavo

djoreyo
I can send you a screen shot of it later but it's definitely 3840x2160. I'm headed out to lunch now.

Yeah, 3840x2160 custom (no AA).

Again, literally not ultra settings.

However, even with custom ultra settings without AA there are serious dips below 60 FPS that break the experience. You and others might not notice it, but I do whenever a game drops below 55FPS - which is why I have to set 1080P ultra as that dips down to 60FPS sometimes, yet not below 55 that I've noticed.

I've noticed that the Wildlands (and many other games) benchmarks are more "stress tests" than representative of actual gameplay.  How often in the game do you make these sweeping panoramic camera moves across dense terrain?  Sub 60's in the benchmarks often translate to 60+ in actual gameplay.   Regardless, Wildlands is not optimized for the 1080ti whereas games like Deus Ex, Hitman, Sniper Elite 4, Tomb Raider, Far Cry 4/Primal and Mafia III are fantastic.  The Division is spectacular!  I'm tempted to reinstall GTA5 but I just kicked that habit a few months ago. 
 
So far the most mind-blowing graphics I've seen on the 1080ti have been on Forza Horizons 3 - both the core game and the new Hot Wheels expansion.  WOW!

In most of the games I've tried with this 1080ti FE card, while I get solid 60fps+ outdoors, inside areas drop to low-30s. Every game I've tried does this, from FO4 to Skyrim and even FO3, and that was with no graphics mods installed. 

trog69
In most of the games I've tried with this 1080ti FE card, while I get solid 60fps+ outdoors, inside areas drop to low-30s. Every game I've tried does this, from FO4 to Skyrim and even FO3, and that was with no graphics mods installed. 

Personally I don't think Bethesda games are really well-optimized. My 1080 Ti SC2 is running it at 4K with max settings and the optional texture pack in Steam and I'm still getting like 50%-60% usage on the GPU while pretty much capping at 60 fps. There are sone areas where the fps dips though.

And no, I don't believe the cpu is the bottleneck when a) I don't have this problem in other games and b) my cpu is an i7-5820K OC'd to 4.4 GHz.

09973
My 1080Ti performance is 100% awesome until heavy physics scenes, performance suffers horridly then. Fire Strike, my physics score is marginally lower than my 780Ti!!!!! OCed my CPU to see if it would help and my physics score dropped a bit lower.

Is that 1080ti installed on the rig shown in your signature?

ThermiteBlaze

trog69
In most of the games I've tried with this 1080ti FE card, while I get solid 60fps+ outdoors, inside areas drop to low-30s. Every game I've tried does this, from FO4 to Skyrim and even FO3, and that was with no graphics mods installed. 

Personally I don't think Bethesda games are really well-optimized. My 1080 Ti SC2 is running it at 4K with max settings and the optional texture pack in Steam and I'm still getting like 50%-60% usage on the GPU while pretty much capping at 60 fps. There are sone areas where the fps dips though.

And no, I don't believe the cpu is the bottleneck when a) I don't have this problem in other games and b) my cpu is an i7-5820K OC'd to 4.4 GHz.

You may be right. I uninstalled Wild Hunt, so I'll have to re-install it and check out how it runs. 
 

arestavo

09973
My 1080Ti performance is 100% awesome until heavy physics scenes, performance suffers horridly then. Fire Strike, my physics score is marginally lower than my 780Ti!!!!! OCed my CPU to see if it would help and my physics score dropped a bit lower.

Physics is 100% taken care of by the CPU, unlike PhysX which can be offloaded to a Nvidia GPU. The rub is that there are precious few games and benchmarks out there that use PhysX (Borderlands series, Batman series, PLA benchmark).

If the firestrike physics score drops when the CPU is OCed, either its an unstable OC or the CPU is getting too hot and downclocking to prevent damage.

 Ah I did not notice that in Futuremark description. I guess I stupidly assumed it was offloaded to the GPU since it was a graphics test. I stressed tested the 4.1GHz OC and it passes a 2 10 minute runs with no hiccups. I just assumed it was a variance since the lost was extremely marginal

I can also confirm I saw a dramatic increase switching between the 1080 ftw hybrid and a 1080 TI SC when running the wildlands ultra benchmark. This is with a I7 7700K. Where I would dip into the 40s during the barrels blowing up stayed consistantly above 65fps with an average over 70fps now. I have also seen a very dramatic increase in VR performance with the vive. I cannot speak for SLI performance though as I don't have money laying around like that haha.

rjohnson11

09973
My 1080Ti performance is 100% awesome until heavy physics scenes, performance suffers horridly then. Fire Strike, my physics score is marginally lower than my 780Ti!!!!! OCed my CPU to see if it would help and my physics score dropped a bit lower.

Is that 1080ti installed on the rig shown in your signature?

Yes, that be the current system, I know its starting to age a bit. RIP on the link=mailto: that I just noticed now lol. Ill update it when I upgrade hopefully this summer. Find a better way to format it next time around.

arestavo

09973

arestavo

09973
My 1080Ti performance is 100% awesome until heavy physics scenes, performance suffers horridly then. Fire Strike, my physics score is marginally lower than my 780Ti!!!!! OCed my CPU to see if it would help and my physics score dropped a bit lower.

Physics is 100% taken care of by the CPU, unlike PhysX which can be offloaded to a Nvidia GPU. The rub is that there are precious few games and benchmarks out there that use PhysX (Borderlands series, Batman series, PLA benchmark).

If the firestrike physics score drops when the CPU is OCed, either its an unstable OC or the CPU is getting too hot and downclocking to prevent damage.

Ah I did not notice that in Futuremark description. I guess I stupidly assumed it was offloaded to the GPU since it was a graphics test. I stressed tested the 4.1GHz OC and it passes a 2 10 minute runs with no hiccups. I just assumed it was a variance since the lost was extremely marginal

If it's a small difference then it's fine - there can be as much as 3% difference between tests.

Well my main wonder about the difference is not between the CPU OCs, but why the physics score was on my 1080Ti is ~8% slower than when I had my 780Ti. Nothing changed except for the Graphics card. (Making Linus happy calling it a Graphics card). Only reason I went to benchmarks is I noticed a good drop in physic performance in games when I went to my 1080Ti. Performance is 100% awesome until a heavy physics scene. 
http://www.3dmark.com/com...s/12712354/fs/12715920