you would think a gtx570 could max any game at 1280...

[AtenRa]

what card do you have ?? cause i thought you had 6950.

 

[WMD]

what card do you have ?? cause i thought you had 6950.

I am running a 5850 with i5 2500k

 

[gorcorps]

Metro 2033 is known to be a very demanding game. GTA4 is known to be one of the most poorly optimized PC games in recent memory. You picked some awful games to use as a test.

You know what it does run? The Witcher 2 without ubersampling. Crysis 2 maxed WITH 3d. Just Cause 2 maxed with 3d. Batman AA maxed with 3d. Deus Ex HR maxed... would be with 3d but its currently broken.

 

[AtenRa]

I am running a 5850 with i5 2500k

ah ok, then yes the dual tess units may do the difference, can you run in higher res ?? say 1080p, i just want to see if the difference in performance will go up at higher res or stay the same.

i will run it later when ill get back home

 

[WMD]

ah ok, then yes the dual tess units may do the difference, can you run in higher res ?? say 1080p, i just want to see if the difference in performance will go up at higher res or stay the same.

i will run it later when ill get back home

I can't. Max 1680x1050 here. But it will be interesting to see the difference between architectures @ same clocks.

 

[toyota]

Hmm not sure why you are having such an issue in GTA IV. I have the same set up as you (2500k, GTX570) and did not run into any issues with that game. I was able to run it at max w/ 4xAA 16xAF and shadows on medium with no noticeable slowdown. As for metro I had to choose between DX11 and medium detail, or DX9 with high detail in order to get decent performance. I didn't notice any real IQ loss either way. It's one of those games like Crysis that wont be able to max out until later generations of hardware.

and how are you running 4x AA in GTA 4?

 

[dpodblood]

and how are you running 4x AA in GTA 4?

Is that not an option in the menu? Maybe I'm mistaken because I don't have it installed anymore. All I remember for sure is that I had it maxed out except for shadows. If 4xAA was an option I would have used it. If not I used whatever the default is.

 

[AtenRa]

I can't. Max 1680x1050 here. But it will be interesting to see the difference between architectures @ same clocks.

I will have to ask toyota first, if it is ok to post the results here in his thread or open a new one

 

[toyota]

Is that not an option in the menu? Maybe I'm mistaken because I don't have it installed anymore. All I remember for sure is that I had it maxed out except for shadows. If 4xAA was an option I would have used it. If not I used whatever the default is.

yeah you are mistaken and GTA 4 is pretty infamous for lacking any AA options. and that game really needs AA too.

 

[toyota]

I will have to ask toyota first, if it is ok to post the results here in his thread or open a new one

fine by me

 

[Dankk]

It's all opinion what games look better than others, and the it's just a console port accusation, debate tactic is getting old. BF3 is also coming out on consoles. There are going to be haters calling BF3 a console port.
Games like NFS Hot Pursuit are console ports, or Star Wars the Force Unleashed.
And there may be people that disagree with me on those ?

I've been playing NFS Hot Pursuit on PC a lot lately, and the graphics are definitely console-ish. Still pretty good-looking, but console-ish. The car models, while slick, almost look too much like plastic. The environments use a lot of samey, low-res textures. It's stuff like this where you can tell the devs had to cut corners so they could accommodate archaic Xbox 360 hardware. The lack of any anti-aliasing options is what bothers me the most.

The "console port" argument doesn't work for everything though. It all depends on how much effort goes into porting it. Dirt 3 is an example of an excellent port from console (rightfully so; Codemasters usually does an good job on making their racing games for PC). Being a console port isn't bad at all in this case. So we have great games like Dirt, but then we also have the Need For Speed games which feel more lazy.

 

[Arkaign]

^^ Fwiw, NFS Shift 2 is dramatically better at reducing the plastic look and cut/paste phenomenon that plagues HP.

 

[notty22]

I love Shift1. Shift2 is pretty good also. Hot Pursuit is the most consolelish, but still very fun to play. Imo, console port, as a dirty word (from pc gamers) is over-used.

 

[Dankk]

^^ Fwiw, NFS Shift 2 is dramatically better at reducing the plastic look and cut/paste phenomenon that plagues HP.

You're probably right. Especially since Hot Pursuit was actually developed by Criterion, and not the regular NFS guys, I can believe it.

To stay on topic: My 6950 1GB makes Metro 2033 quite playable on DX11 "Very High" settings at 1080p (nowhere near 60 fps though, oftentimes dipping into the teens... but still "playable"). The 6950 is about neck-and-neck with the GTX 560 Ti, so I'm not sure why a 570 wouldn't be able to do it. Even more puzzling is that it struggles at a mere 1280x960.

I would say maybe it's CPU-related, but the OP is running the same processor as me (i5 2500k) so that's probably not the issue.

Maybe it has to do with the benchmark being significantly more demanding than the rest of the game. But even then, the framerate dropping into the teens for so long still doesn't sound right. Something about the OP's computer must be gimped.

I'll try running the benchmark when I get home from work.

 

[toyota]

its certainly the Metro2033 benchmark that is demanding in two spots as the actual game is fine from what I can tell.

 

[AtenRa]

AvP

1680 x 1050/Very High/High/16/SSAO Off/TS On/AS On/0X
Used AvP_D3D11_Benchmark.exe

Number of frames: 7757
Average Frame Time: 13.5ms
Average FPS: 74.0


1680 x 1050/Very High/High/16/SSAO On/TS On/AS On/0X
Used AvP_D3D11_Benchmark.exe

Number of frames: 7304
Average Frame Time: 14.4ms
Average FPS: 69.6fps

Difference is ~6,32%

 

[cusideabelincoln]

Xbox 360 GPU ~ X1900XT = 7900GTX

I don't think these two are quite so equal in newer games and console ports.

http://www.tomshardware.com/charts/.../compare,746.html?prod[2143]=on&prod[2093]=on

The 7900 is no match for the X1900 in console ports Call of Duty 4, Mass Effect, and Asassin's Creed. The X1900 is almost twice as fast in CoD.

The X1900 is also appreciably faster in Crysis and HL2: E2, but perhaps not by an abnormal amount.

In the non-console-ports (FS, WiC, ETQW) they either tie or the 7900 is a bit faster.

 

[WMD]

:thumbsup:

In this game, the DX11 performance hit is enormous.

http://gamegpu.ru/Action-/-FPS-/-TPS/Metro-2033-Direct-X-11-test-GPU.html

Very HQ 1280x1024 AAA/16x AF in DX10: HD5870 = 48 fps

vs.

Very HQ 1280x1024 AAA/16x AF in DX11: HD5870 = 31 fps

This is one of the problems with DX11 implementations in current games. In order to "showcase its feature the developer with the help of hardware vendors push alot of stuff to run on DX11 exclusive compute shaders. The performance benefit claims of running these graphical effects using compute shader is questionable. Even "bokeh" type DOF effects can be coded to run on regular pixel shaders with only moderate perfomance hit as demonstrated by Witcher 2. In game FXAA and MLAA options in recent games have also shown they both performs much better than AMD compute shader based MLAA.
The very fact that games are developed with "help" from hardware vendors to implement certain new graphical feature should be bad news from a performance perspective. Graphic hardware vendors have all the incentive to optimize games for poor performance rather than the opposite.

 

[WMD]

AvP

1680 x 1050/Very High/High/16/SSAO Off/TS On/AS On/0X
Used AvP_D3D11_Benchmark.exe

Number of frames: 7757
Average Frame Time: 13.5ms
Average FPS: 74.0


1680 x 1050/Very High/High/16/SSAO On/TS On/AS On/0X
Used AvP_D3D11_Benchmark.exe

Number of frames: 7304
Average Frame Time: 14.4ms
Average FPS: 69.6fps

Difference is ~6,32%

Here's mine:

SSAO on

Benchmark Summary:

Number of frames: 6734
Average Frame Time: 15.6ms
Average FPS: 64.2

 

[BathroomFeeling]

Dunno if this is helpful, but here are my stats for my non-OCed system:

AvP with nVidia 266.58 drivers:
1680 x 1050/Very High/High/16/SSAO Off/TS On/AS On/0X
AvP_D3D11_Benchmark.exe 1.03
Number of frames: 8021
Average Frame Time: 13.1ms
Average FPS: 76.5


1680 x 1050/Very High/High/16/SSAO On/TS On/AS On/0X
AvP_D3D11_Benchmark.exe 1.03
Number of frames: 7489
Average Frame Time: 14.0ms
Average FPS: 71.4fps

6.667% reduction.




AvP with nVidia 280.26 drivers:
1680 x 1050/Very High/High/16/SSAO Off/TS On/AS On/0X
AvP_D3D11_Benchmark.exe 1.03
Number of frames: 7281
Average Frame Time: 14.4ms
Average FPS: 69.5


1680 x 1050/Very High/High/16/SSAO On/TS On/AS On/0X
AvP_D3D11_Benchmark.exe 1.03
Number of frames: 6285
Average Frame Time: 15.4ms
Average FPS: 65.1fps

6.331% reduction.

 

[DirkGently1]

Define 'max'? In-game settings? Custom configs? Driver level settings? Third party utilities?

Apply enough SGSSAA, for example, to a game and any rig will fall to it's knees at that res.

 

[toyota]

Define 'max'? In-game settings? Custom configs? Driver level settings? Third party utilities?

Apply enough SGSSAA, for example, to a game and any rig will fall to it's knees at that res.

I was talking about the max settings available from within the game.

 

[AtenRa]

Here's mine:

SSAO on

Benchmark Summary:

Number of frames: 6734
Average Frame Time: 15.6ms
Average FPS: 64.2

1280x960

6950 84.5
5850 80.3

Difference 5.23%

1680x1050

6950 69.6
5850 64.2
Difference = 8.41%

Can you tell me the memory clock of your card?? mine is at default 1250MHz.

Memory bandwidth may play a bigger role than tess in higher ress in this bench.

 

[trollolo]

Mme. GTA4 is known to be one of the most poorly optimized PC games i

has the GTA been patched yet?

 

[Dankk]

has the GTA been patched yet?

GTA4 has been patched numerous times since release, however it's been a year or two now and it's probably as patched as it's going to get. It runs far better than it did on release, but it's still a poorly optimized game.