Stone Giant DX11 Demo

Hey  
Thanks,will download and post my results as soon as its done(with a 36KB/S DL speed  )

Great,it doesn't have resume support ?
Never mind,here is a download link from techpowerup and supports resume
http://www.techpowerup.com/downloads/1791/Stone_Giant_D...

Unfortunately I cant edit my posts.

Here is results with OC (920/1150):

ALL MAX - 26
DoF off - 37
Tess Med - 54
Tess Off - 64

Thanks for posting rawsteel, I'll run the benchmark and post back with results  

I am interested to see GTX4-- cards benchmarks. It will probably greatly favor them with the high tesselation levels.

Anyone with Fermi to post?

Thanks in advance

DoF off for all Runs

Tesselation High at 1920 x 1080 = 39FPS Avg

Tesselation on Med @ 1920 x 1080 = 58FPS Avg

Tesselation off @ 1920 x 1080 = 84FPS Avg

Spec in sig

1680x1050, Tesselation High, DOF On

GTX 470 stock - 37

GTX 470 OCed 703/1405/1755 - 42

DOF OFF:
Tesselation Off
stock - 82
OC - 93

Tesselation Med
stock - 73
OC - 82

Tesselation High
stock - 64
OC - 72

wow omgitzfatal, how did you get such results?

Can you run at 1680x1050?

Wow, Fermi's really shine on Tesselation. Thats great!

What I noticed is how little frames drop for you between High tesselation and off 82/64 - 28%
While for me and ATI 54/30 - 80%

Lets see DoF : You - 64/37 - 73%
Me - 30/22 - 36%


Still 470 is 70% faster than 5850 (stock) on this benchmark - impressive


May be we should not sign nVidia off yet  

I've had 2 cards in hand for 2 days before deciding which one stays: HD 5870 and GTX 470. That's pretty much why I decided on Fermi: at my native res it seemed like a better deal than 5870, considering its DX11 performance. HD 5870 is a way to go though if you are on 1920x1080 and up and dont care about physx/cuda, no questions there.

Great,when i click on it it starts in windowed mode(with windowed mode unchecked) and as soon as i go to full screen,it becomes slow and crashes after few seconds  

Is there a benchmark button anywhere?

Hit Enter, make sure you toggle F4 to see results.

You benchmark by hitting Enter when the demo loads

I guess it can be buggy, Maziar whats your VC?

F1 - toggle wireframe
F2 - toggle Tesselation
F3- toggle DoF
F4- toggle FPS

1-5 buttons - switch several static views

Space - toggle mouse view
wasd - navigate

Tesselation HIGH, 1440x900 , average 43 fps


on :
5870
i5 2,66

Ya I found it by reading the readme. Who actually reads those lol.

With everything at 1680x1050 I scored 144FPS

Using i7 920 @ 3.8ghz and a 5970.

You sure you ran it with DOF? Its off by default.

Where i can change DOF and others details? I see only tesselation

Oh no I didn't.

That would be 128FPS with Depth of Field

Rofl thats impossible

its more than 4x my score

You must be doing something wrong

Check if you turn on DoF ( its off by default) (F3)

are you kidding me - thats simply impossible

Do you have quad crossfire 2x5970?

Great results, awesome DOF scaling.

Maybe your processor is bottlenecking? Im also running it fullscreen.

What do you mean by VC? Vcore ?
Anyway it runs fine when in full screen when i disable CrossFire but with CF enabled it runs extremely slow in full screen

Starting it in fullscreen doesn't work with crossfire for me either.

Start it in windowed then hit ALT+ENTER which would be the universal hotkey for fullscreen.


I think crossfire might be a bit buggy. In the benchmark the entrance of the cave gives me 9 FPS. If that scene was taken out my bench would be several frames higher.

Demo is buggy. I cant change resolution at all  

It runs only 1680x1050 no matter what I chose from the dropdown

By VC I mean Video Card  

Also enabling 4xAA from catalyst makes it run 2fps FASTER! but looks MUCH MORE blurry

btw I think it have some sort of AA as it looks amazing without any AA set (may be UT3 engine like or something)

Did you look at the wireframe on that? There were so many edges everywhere. I wouldn't be surprised if the graphics card mistook some of the edges as hard edges and applied AA to make it burry.

AA is just not necessary, there are so many edges to smooth things over. Not the mention the demo was dark. It would be hard to notice hard edges on a black background.

It looks like this demo applies some sort of AA/AF combination on nvidia cards even if there is no options for it in the demo itself:

I'm getting avg 40 FPS with AA/AF truly disabled in nvidia control panel on stock @ 1680x1050, high tesselation, DOF on and 74 FPS with DOF off as opposed to 37 FPS and 64 FPS with "let the application decide" options for AA/AF selected in control panel.

I'm going to repost all my runs with AA/AF truly disabled.

How come on such low resolution your card be 4x mine? There is something iffy here

Even on perfect scaling you can expect no more than 2x. Except if my card is severely bottlenecked at something.

Its not cpu, I tried overclocking more and it didnt up the fps even with 1

also you are getting MUCH higher results even than 480GTX which you shouldnt

What driver are you using?

I tried to run in Windowed mode and the ALT+Enter but it crashes after loading  

Im using 10.4 drivers and why shouldn't I get more than a GTX 480?

Because as in Unigine GTX480 is about same or little faster on High tesselations and I've seen on the forums guys with 480GTX get something like 47-50 fps on Tesselation High + DoF

No way you can get 128 fps, just something is wrong obviously, either not applying tesselation/DOF or fps metter is broken

Try measure with Fraps see if you get similar results

i got 47 avg with everything maxed on 1280x1024 with a 5850 1GB 950/1250  

GTX 470, 1680x1050, AA/AF disabled in control panel:

Stock, DoF OFF
Tesselation off - 96
Tesselation med - 85
Tesselation High - 72

Stock, DoF ON
Tesselation Off - 46
Tesselation Med - 43
Tesselation High - 40

OCed 703/1405/1755, DoF OFF
Tesselation Off - 108
Tesselation Med - 96
Tesselation High - 82

OCed 703/1405/1755, DoF ON
Tesselation Off - 51
Tesselation Med - 48
Tesselation High - 44

I ran it 3 more times and DoF was definitely on and tesselation is definitely high.

Run 1: 84FPS
Run 2: 128FPS
Run 3: 128FPS

I think the first run wasn't that good because the card needed time to get off idle speeds. So if you have an ATI card, you might want to run it more than once to get your card warmed up.

I ran the demo a few different times, with very different results on:
i7 920 @ 3.3 and 3.5 with 5870's in CF at a resolution of 1920x1200.

In windowed mode with cat 10.3 drivers and everything on and high tess: 47

In windowed mode with cat 10.4 drivers and everything on and high tess: 49
In fullscreen mode with cat 10.4 drivers and everything on and high tess: 119 (at least this is what is says).

When I traveled into the little tunnel, not doing the benchmark, it would drop into the 30's.

Sounds like something is bugged with windowed mode.

Windowed mode would always tax performance. Even though nothing is happening in the desktop, the position of the windows still needs to be continously updated and redrawn. More taxing if you have aero turned on.

Another weird thing I noticed is that when I'm controling the camera, and moving around the cave under my control, I got lower FPS with crossfire than with a single 5870, but crossfire did increase performance on the benchmark.

Edit: I also noticed another oddity on the fullscreen mode benchmark that gave the 119 average FPS result. During the whole benchmark, I was seeing numbers over 200-300 FPS while fluttering with much lower numbers. I should have seen an average FPS over 200 by the way it looked, but it wasn't, meaning that multiple times a second, it showed FPS over 300 and below 100.

Here is my score:

1440x900 , tesselation ON , DOF OFF:

average: 43 FPS



Uploaded with ImageShack.us


1440x900, tesselation ON , DOF ON:

average: 30 FPS



Uploaded with ImageShack.us

little edit: i mean all TESSELATION on HIGH


But why someone write a few posts above , that he had on 5850 average fps
47 on MAX, better than me on 5870 ? Lier? i got like i say 30 fps on ALL MAX
DOF ON , TESSE HIGH

He could have a lower resolution. He could also have it OC'ed to 5870 speeds. It's also possible he didn't turn on DoF, as it's something you have to turn on in the demo, and he didn't realize it.

At 2560x1600 with high tesselation. DOF off because it locks up at that high of resolution.

Scored 107 FPS in that bench.

Looks nice... What are your specs?

God i just love dx11 effects. Makes me want to replace my GTX 295 with a 480 (cause of the tesselation performance). But it still performs like a beast so i'll probably just wait till the ATI 6xxx or refresh series  

Im not a liar   my 5850 is overclocked to 950 core and 1250 memory and my resolution is only 1280x1024. and i had everything turned to max, DOF=ON and Tesselation=HIGH 

He didn't call you a liar. He said lier, completely different  

What CPUs are you guys running. My 5970 is getting not only twice the frames but 3-4x the frames on single 5870 and 5850s.

Of course my 5970 is overclocked to 950/1250

I think it have something to do with Catalyst driver version.

All using 10.3 are getting somewhat consistent fps and those using 10.4 some super high unnormal fps. May be its driver issue not applying something or just showing wrong fps (most probably)