[Gamegpu.ru]Skyrim: Special Edition GPU benches

[Jaydip]

http://gamegpu.com/rpg/роллевые/the-elder-scrolls-v-skyrim-special-edition-test-gpu

Running pretty well across the board.

 

[TheRyuu]

Yet another benchmark without frame time analysis!

 

[KompuKare]

Yet another benchmark without frame time analysis!

Well, the Computerbase Skyrim Special Edition article does have frametimes:
https://www.computerbase.de/2016-10/skyrim-special-edition-benchmark/3/#abschnitt_frametimes
Unfortunately, only for two cards though.

 

[raghu78]

Nvidia cards perform very well in this game as expected. AMD cards are significantly slower than their Maxwell or Pascal counterparts. The GTX 1060 has almost a 20% lead over Rx 480.

 

[Krteq]

Yep, CrapWorks™ :/


Trolling is not allowed
Markfw900
Anandtech Moderator

 

[arandomguy]

What's more interesting is the CPU benches at gamegpu. Would be interested in some more tests to see how much mins and consistency improved in worst case scenarios. I don't believe they tested in one of the worst areas, maybe starting a mass fight in certain areas of Markarth or Riften?

 

[Face2Face]

Gopher did a frame rate comparison on his YouTube channel and the SE runs better vs. the original, while maintaining better visuals.

https://www.youtube.com/watch?v=JRzLgRvqfDs

The Godrays in this game are not nearly as demanding as FO4's, but don't look quite as good either. I will say looking up at an overhead dragon with the godrays coming through is quite amazing.

 

[Bacon1]

Well, the Computerbase Skyrim Special Edition article does have frametimes:
https://www.computerbase.de/2016-10/skyrim-special-edition-benchmark/3/#abschnitt_frametimes
Unfortunately, only for two cards though.

Really shows how bad AMD is CPU limited in this game.

1440p narrows the gap between the cards massively.

1080: Fury -> 980 Ti = 18%
1440: Fury -> 980 Ti = 10%

Same with 480 -> 1060, 18% -> 13%

Makes sense though when the CPU scaling looks like this:

 

[NTMBK]

Really shows how bad AMD is CPU limited in this game.

1440p narrows the gap between the cards massively.

1080: Fury -> 980 Ti = 18%
1440: Fury -> 980 Ti = 10%

Same with 480 -> 1060, 18% -> 13%

Makes sense though when the CPU scaling looks like this:

That graph says to me that all of those processors run that game just fine (apart from maybe the FX-4100 and 6100). Not really seeing the problem.

 

[USER8000]

The physics in the game is linked to framerates since it uses the same engine as Fallout 4 ,so review sites need to artificially remove the limitation. Above 60FPS it can cause issues.

I tried it on my GTX960 and it does seem to run fine,and has better visuals,although the original game with mods will look better.

So,hopefully once more mods for Skyrim MK2 come out,we might get some reviews(ENB based ones will be a good start when they are released).

That graph says to me that all of those processors run that game just fine (apart from maybe the FX-4100 and 6100). Not really seeing the problem.

Not sure if it is that much of a problem - even the slow FX4100 is hitting 60FPS in the game.

You are going to more graphics card limited - the original was worse since it used an X87 path which modders initially had to patch to a faster one(Bethesda did an official patch at a later date).

 

[frozentundra123456]

That graph says to me that all of those processors run that game just fine (apart from maybe the FX-4100 and 6100). Not really seeing the problem.

You mean the 2100 (SB) not the skylake 6100. In any case, seems a lot less cpu demanding than FO4. It actually looks better and runs better than the original on my low end system. Only uses about 20% cpu with my HD 7770. Makes me seriously consider a gpu upgrade, but hopefully they will bring out a new single player game on this engine one of these days.

 

[AdamK47]

Where are the 6950X and Titan Xp results?

 

[USER8000]

You mean the 2100 (SB) not the skylake 6100. In any case, seems a lot less cpu demanding than FO4. It actually looks better and runs better than the original on my low end system. Only uses about 20% cpu with my HD 7770. Makes me seriously consider a gpu upgrade, but hopefully they will bring out a new single player game on this engine one of these days.

Give it some time - with Fallout4,some of the modders made texture mods and other mods for slower systems,which improved performance quite a bit. Apparently,Skyrim MK2 uses the same engine.

 

[Bacon1]

That graph says to me that all of those processors run that game just fine (apart from maybe the FX-4100 and 6100). Not really seeing the problem.

I'm not saying it had bad CPU usage, what I meant was because every single CPU was slightly better than the last, that the game isn't hitting a GPU bottleneck at all at 1080p and that it is hitting a CPU bottleneck..

Since AMD has more overhead on DX11 they'll always be behind slightly when CPU limited.

 

[Head1985]

What's more interesting is the CPU benches at gamegpu. Would be interested in some more tests to see how much mins and consistency improved in worst case scenarios. I don't believe they tested in one of the worst areas, maybe starting a mass fight in certain areas of Markarth or Riften?

Far more better.I have instaled 200+ mods and i still have rock solid 60fps.This new dx11+64bit engine is best thing for moders.
In old dx9 engine i get bellow 30fps with mods like JK cities.Now i have 60fps everytime.

Skyrim dx9-you can see 48% GPU usage and 26fps-Its cpu bottleneck.
https://abload.de/img/tesv_2016_10_29_07_22r2zi2.png
SE edition same mods and 60fps(no ENB)
https://abload.de/img/skyrimse_2016_10_31_052zmb.png

 

[Headfoot]

You guys are making me want to reinstall modded out Skyrim SE again now....

 

[NTMBK]

Oh

You mean the 2100 (SB) not the skylake 6100. In any case, seems a lot less cpu demanding than FO4. It actually looks better and runs better than the original on my low end system. Only uses about 20% cpu with my HD 7770. Makes me seriously consider a gpu upgrade, but hopefully they will bring out a new single player game on this engine one of these days.

I actually meant the FX 6100, the i3 looks fine to me

 

[Grooveriding]

 

[Face2Face]

Where are the 6950X and Titan Xp results?

Pick A resolution, any resolution and your setup will hold 60 fps no problem

 

[[DHT]Osiris]

Not sure what's wrong with their benchmark.. they're citing 49FPS avg on a 970 @1440p, I'm currently running at a constant 60FPS @1440p (vsync, so real numbers are probably a lot higher), everything on, ultra'd, and with graphical mods. I am running a 6700k (oc'd) but the CPU is sitting at like 5-10% in-game so kinda feel like it shouldn't be getting restricted in any way from a CPU standpoint.

Am I missing something?

 

[Head1985]

Not sure what's wrong with their benchmark.. they're citing 49FPS avg on a 970 @1440p, I'm currently running at a constant 60FPS @1440p (vsync, so real numbers are probably a lot higher), everything on, ultra'd, and with graphical mods. I am running a 6700k (oc'd) but the CPU is sitting at like 5-10% in-game so kinda feel like it shouldn't be getting restricted in any way from a CPU standpoint.

Am I missing something?

GTX970 1550Mhz
https://www.youtube.com/watch?v=YnlzIjF4weA

 

[[DHT]Osiris]

I need to fire it up at the start to mirror that then. Currently using an alt-start mod so I didn't go through the intro scene. What gives that overlay? I've been trusting my monitor's OSD which may not be accurate with gsync off.

 

[Bacon1]

I've been trusting my monitor's OSD which may not be accurate with gsync off.

If its like my monitor, then it will display the panel refresh rate, not your FPS. So thats probably why it is @ solid 60 while in game fps might be lower / higher (if vsync was off).

Most people use MSI Afterburner (don't need msi brand card) for overlay, or steam overlay, or DXtory for UWP / DX12 apps.

 

[[DHT]Osiris]

Solid, I'll check out afterburner when I get home.

Yeah, I use gsync for basically everything so I've trusted that OSD in the past, but i'm sure you're right.

 

[KompuKare]

PCGH's benches are out too:
http://www.pcgameshardware.de/TES-5...Spiel-57338/Specials/Benchmarks-Test-1211808/
Kepler is doing very well there indeed. Guess the engine is not compute heavy or something. But then the question is, how well optimised for the consoles is it?